Neem (Azadirachta indica)

In brief

The tree, and its place in two medical systems

Neem is a tree of the mahogany family, Meliaceae, an evergreen that grows 20 to 23 meters tall with a straight trunk, compound leaves made of 5 to 15 leaflets, and small green fruits, botanically drupes, that turn golden yellow when they ripen between June and August ¹. It is native to the Indian subcontinent and is now planted across the tropics, from the Sahel to Southeast Asia, and the poisonings reported from Malaysia, Singapore and Sri Lanka as well as from India occurred in South Asian families who gave the oil to infants as a home remedy ^1,22,33,34. Its genome was published in 2012, the first for any member of the family: rich in two of the four letters of the genetic code, adenine and thymine; poor in the repetitive DNA that bulks out many plant genomes; about 20,000 genes; and grouped, by comparison of gene sequences across species, with the sweet orange, which confirmed the old placement of the mahoganies next to the citrus family ³. A 2016 reassembly with long sequencing reads tripled the length of the unbroken stretches of assembled sequence, cut the number of places where the assembly had been stitched together wrongly 13.4-fold, and mapped the genes of terpenoid synthesis better ². The first enzymes on the road to azadirachtin were identified in 2019: a ring-closing enzyme from neem that folds the 30-carbon starting chain into tirucalla-7,24-dien-3beta-ol, and two cytochrome P450 enzymes, from chinaberry and from sweet orange, whose three oxidations turn it into the early limonoid melianol ⁴. A 2023 study then reconstituted 22 enzymes from chinaberry and citrus into complete routes to two furan-bearing limonoids, azadirone and kihadalactone A ⁵⁷, and a 2026 study mapped the same family of ring-closing enzymes in a third Meliaceae tree ⁵⁸. The many later steps that would turn such a limonoid into azadirachtin itself are still unknown, which is why the pesticide is still extracted from seed ^4,57.

Its medical history is older than its medical literature. Kumar and Navaratnam traced the tree's use from palm-leaf manuscripts through the Ayurvedic texts to the present, under headings that include skin disease, smallpox, diabetes, gut disease, high blood pressure and urinary disease, and noted that the uses predate 'the availability of written records which recorded the beginning of history' ⁵. The tree sits in Ayurveda, in Unani, the Greco-Arabic system practiced in South Asia, in Siddha and in homeopathy, and every part of it, leaf, bark, seed, flower, fruit, root and twig, has a use ^1,6,7. A Unani institute in Bengaluru still formulates it, for instance into a turmeric-neem-Cassia soap tested against ringworm in a trial published in 2022 ⁵⁹. The seed oil, called margosa oil in older English and kohomba oil in Sinhala, is 'used mainly for external applications' but 'is often administered orally to neonates and infants regularly in small amounts', in the words of the Singapore pediatricians who treated two of them ³³; in Sri Lanka it was given to an infant as a home remedy for intestinal worms ³⁴.

What is in it is better known than what it does. More than 140 compounds have been isolated from the tree ^6,8. The characteristic ones are limonoids, a family of highly oxygenated terpenoids: a triterpene has 30 carbons, a limonoid has lost four of them, which is why the chemists call it a tetranortriterpenoid, and has closed a furan ring, a five-membered ring with one oxygen ^9,15. The azadirachtins are limonoids with the C ring broken open; nimbin and salannin belong to neighboring classes with the same broken ring; gedunin has the D ring opened instead; nimbolide is a salannin-class molecule; and nimbidin is not a single compound but a mixture of seed tetranortriterpenes ^15,60. Azadirachtin is concentrated in the seed kernel: a standardized commercial extract of kernels contains 34% azadirachtin A and 56% limonoids in all ⁶¹, while the pressed oil contains 300 to over 2,500 parts per million of it, the rest being triglycerides and plant sterols ¹⁰.

The tree's reputation outruns its evidence, and the gap is visible in its own literature. A 2025 review is titled 'A Miracle Herb; Panacea for All Ailments' ⁶²; a 2016 review ends that the tree 'modulates the numerous biological processes without any adverse effect' three sections after it has summarized the infant poisonings ¹. A 2026 narrative review of the literature to November 2025 counts only fifteen registered interventional trials of neem, most for conditions of the mouth, with 'no published phase II/III results for systemic diseases', meaning results from the mid-stage and late-stage trials that test whether a treatment works, and that the toxicology confirms 'pediatric encephalopathy from neem oil ingestion and embryotoxicity' ⁸. That sentence is the frame for everything that follows.

The whole history, including the reversals

The history runs on two tracks that rarely meet: a chemistry track, in which an insect antifeedant was isolated from neem seed in 1968 and the world spent the next 39 years working out what it was and how to make it, and a medical track, in which the same seed's oil was found in 1981 to be killing infants, and in which almost every therapeutic claim was tested in a dish or an animal and almost none in a trial ^8,11,12,22.

On the chemistry track the dates are secure. Azadirachtin was isolated from neem seed in 1968 by J. H. Butterworth and E. D. Morgan at Keele University, as the substance that stops locusts feeding, a history recorded by Morgan himself and by the Cambridge review rather than in any PubMed-indexed original ^11,12. A structure was published in 1975 by Zanno, Miura, Nakanishi and Elder, worked out by carbon-13 nuclear magnetic resonance, a method that maps a molecule's carbon atoms by how they respond in a strong magnetic field ³²; it did not stand. E. David Morgan, one of the discoverers, writes that its structure 'required 18 years to solve' and its synthesis 'another 22 years', which places the settled structure in the mid-1980s, nearly a decade after the first proposal, and the synthesis in 2007 ^11,12. The Cambridge group of Steven Ley published the first total synthesis in two back-to-back communications in 2007, and the full account the following year carried 46 authors, the roll of two decades of students and postdoctoral chemists ^13,14,84. By then more than 1,000 papers had been written about the molecule ¹², and by 2021 more than 2,000 ¹⁵. Morgan's 2009 reflection, 'Azadirachtin, a scientific gold mine', is also candid about what was not known: the mode of action, the structure-activity relationships and the biosynthesis 'still require much research', and the pesticide had 'still not achieved a prominent place' and was unlicensed in many countries 40 years after its discovery ¹¹.

The medical track begins with a reversal. In February 1981 Sinniah and Baskaran reported in the Lancet thirteen Malaysian infants who developed vomiting, drowsiness, metabolic acidosis, neutrophil leukocytosis and encephalopathy within hours of being given margosa oil, with fatty liver and kidney and mitochondrial damage on biopsy and in experimentally poisoned mice ²². An Indian series the following year, summarized by later authors as 12 children given 25 to 60 ml of whom 10 died, confirmed it ^10,23. The oil had been given because it was believed harmless; the belief was wrong, and the series that overturned it were followed over four decades by a rat model, a mitochondrial mechanism, a mouse time course, a treatment study and a steady trickle of cases from Singapore, Sri Lanka and India, the last in 2014 ^{33,34,35,70,72,73,85}.

The therapeutic claims were tested one program at a time. The National Institute of Immunology in New Delhi spent the 1990s on contraception: a single intrauterine dose of neem oil blocked rat and monkey fertility for months, reversibly, and the active fraction turned out to be free fatty acids rather than any limonoid ^86,87,88; the same group's Praneem formulations went through phase I and II safety trials as vaginal microbicides in the 2000s and stopped there ^89,90,91. Malaria was tested in dishes from 1985, when nimbolide stopped Plasmodium falciparum in culture at about 2 micromolar while neither it nor crude neem extracts did anything to Plasmodium berghei in infected mice, by mouth or by injection ⁹²; gedunin was identified between 1986 and 1989 as the leaf limonoid responsible, about as potent as quinine in culture ^93,94; and the work has gone on in dishes and mice ever since without reaching a patient ^39,40,61,95. The mouth was tested most, in a run of small chlorhexidine-controlled trials from 2004 that a 2016 systematic review found 'generally flawed' ^30,96. The 2020s added a pilot COVID-19 prevention trial with eleven events ⁹⁷, an 80-person metabolic-syndrome trial ⁹⁸ and a 61-student scabies trial ⁹⁹.

The patent fight ran in parallel. A European patent on a fungicidal neem-seed preparation, number 0436257, was revoked by the European Patent Office in 2000 after an opposition by environmental and organic-farming groups, and in March 2005 the Technical Board of Appeal left the revocation standing, finding the main claims obvious to try and the fallback claims to contain added matter ^25,26. The case became the textbook instance of what activists called biopiracy, and a 2006 bioethics analysis argued that the allegation deliberately simplified both the science and the law in order to attack the patent system itself ²⁷. In the United States the first neem product had been registered fifteen years earlier, in 1985, as Margosan-O, a seed-extract concentrate standardized to 3,000 parts per million of azadirachtin, for nonfood uses ²⁴.

Azadirachtin: the molecule, with the other limonoids beside it

Azadirachtin has the formula C35H44O16 ¹⁰. From that formula its molar mass works out to about 720.7 grams per mole, which is to say that one molecule carries 35 carbons, 44 hydrogens and 16 oxygens, an oxygen for every two carbons, which is a great deal of oxygen for a terpenoid and part of why the structure was so hard to solve ^10,11. The molecule is a limonoid of the ring-C-seco class, meaning the third ring of the triterpene skeleton has been opened, and it is one of a family: fifteen azadirachtins are known, lettered between A and Q, with the letters C and J unused; ten of them carry azadirachtin's own skeleton and five, 11alpha-H, I, L, M and P, have lost a fifth carbon, and A is the one usually called azadirachtin, the commercial extracts containing A at 34% with the others adding 16% more ^15,61. It is a feeding deterrent for many insects and a growth disruptant for most, and for many other arthropods ¹¹; one review puts its activity at more than 400 insect species, with antifeeding and growth inhibition 'the most remarkable' of its effects ¹⁵.

The synthesis is the part of the story chemists tell. Ley's route builds two halves separately, a pair of fused six-carbon rings and a second fragment carrying a reactive triple bond, and joins them through an oxygen atom. A Claisen rearrangement, a reaction that turns a carbon-oxygen link into a new carbon-carbon bond, then forges the crowded bond between carbons 8 and 14. A ring closure driven by an unpaired electron, which chemists call a radical cyclization, makes the bridged two-ring cage of the natural product, and 'a series of carefully designed steps' that drew on the group's earlier study of how azadirachtin itself reacts finishes the molecule ¹⁴. The two 2007 communications are titled 'a long but successful journey' and 'a relay route', the second completing the molecule from an intermediate obtained by degrading natural azadirachtin ^13,84. The synthesis has not displaced the tree: azadirachtin for pesticides and research still comes from seed, and when a research group bought it commercially in 2010 the bottle held about half the declared amount ⁶¹.

The other limonoids are less famous and better studied in mammals. Nimbidin, the major active principle of the seed oil, is a mixture rather than a compound; in rats it suppressed the migration, the phagocytosis, meaning the engulfing of particles, and the respiratory burst, meaning the blast of oxidants immune cells use to kill microbes, of macrophages at 5 to 25 mg/kg by mouth, and in a dish cut those cells' output of nitric oxide and of prostaglandin E2, a signaling molecule of inflammation, and the degranulation, or release of stored enzymes, of neutrophils ⁶⁰; its anti-arthritic activity in rats was reported in 1981 ⁷⁴. Nimbin, salannin and 6-desacetylnimbin, like azadirachtin, inhibit the insect ecdysone-activating enzyme, though at concentrations of 2 x 10^-5 M and above ⁶³. Gedunin, a leaf limonoid with the D ring opened, was the antimalarial principle of neem leaf in a 1997 screen of 22 Meliaceae, the only compound of the set to beat chloroquine, the old standard antimalarial drug, against the parasite clone called W2; that paper labels W2 chloroquine-sensitive and D6 chloroquine-resistant, which reverses the standard designations of those two clones, so the result is reported here as its authors stated it ³⁹; in 2006 a chemical-genomic screen for inhibitors of androgen-receptor signaling, the male-hormone pathway that drives prostate cancer, found it and predicted from its gene-expression signature, then confirmed, that it inhibits heat-shock protein 90, the chaperone that many cancer-driving proteins need to fold ⁷⁵. Its relative deoxygedunin, given to mice, activated TrkB, the receptor for brain-derived neurotrophic factor, a nerve-growth signal, and prevented a nerve degeneration in pups lacking that factor ⁷⁶. Nimbolide is 'the most potent cytotoxic neem limonoid' in cancer cell lines and animal models ⁷⁷; in 2025 and 2026 alone it was reported to protect mouse liver from acetaminophen when given two hours before the poison ⁷⁹, to shrink aortic plaque in mice by binding a 14-3-3 protein, one of a family of adaptor proteins that hold other signaling proteins in place ⁷⁸, and to be, in a 2025 review's words, 'still in the early stages of the drug development process because no comprehensive pharmacokinetic studies or long-term toxicity studies' exist, pharmacokinetic meaning the study of how the body absorbs, distributes, breaks down and excretes a drug ⁸⁰. The 2024 review that assembles the cancer work for all four reports no human trial of any of them ⁷⁷.

How it works on insects, and why not on us

An insect that meets azadirachtin stops eating. The primary antifeedant effect is mediated by the gustatory chemosensilla, the taste hairs of the mouthparts: the molecule activates bitter-sensitive cells and inhibits the firing of sugar-sensitive ones, and a secondary, slower loss of appetite follows from damage to gut, muscle and fat body ¹⁵. Insect taste preparations respond at 1 to 100 nM ²¹. The effect was the one that got the molecule found: it was isolated in 1968 as an insect antifeedant ¹², and the desert locust is the insect in which its binding sites were later mapped ⁶⁷.

An insect that eats anyway stops growing. Azadirachtin is an antagonist of 20-hydroxyecdysone and of juvenile hormone, the two hormones that govern molting and metamorphosis, and its major action is on the titers of both: it inhibits release of prothoracicotropic hormone from the brain complex that would otherwise instruct the prothoracic glands to make ecdysone, it inhibits ecdysone 20-monooxygenase, the enzyme that converts ecdysone to the active 20-hydroxy form, and it produces degenerative changes in the nuclei of all three endocrine glands that control the molt ¹⁵. The enzyme inhibition was measured in 1988 in homogenates of fruit fly, mosquito and tobacco hornworm, with half-inhibition at 1 x 10^-4 to 4 x 10^-4 M ¹⁷, and again in 1997 for nimbin and salannin as well ⁶³; those concentrations are far above the nanomolar ones that act on whole insects, so the enzyme is not the whole story. The experiment that shows the hormone is the point was done in the kissing bug Rhodnius: azadirachtin inhibited molting and reproduction, ecdysteroid titers fell, and the molting block 'is fully reversed by ecdysone therapy' ¹⁶. The same rescue works in sand-fly larvae, where adding ecdysone to the diet counteracted both the mortality and the block of molting ¹¹². One 2013 mosquito-egg paper filed azadirachtin as an 'ecdysone agonist' in its abstract ¹¹³; the insect-physiology literature and the rescue experiments treat it as an antagonist of the molting hormone ^15,16,112, and the label is best read as a slip, although on a different target, the dopamine-releasing nerve cells, one study did report an agonist-like effect ¹⁵.

Where it binds is only partly known. Dihydroazadirachtin labeled with tritium, a radioactive form of hydrogen that lets the bound molecule be counted, binds a single population of sites in insect cell nuclei with a dissociation constant of 18.1 nM, essentially irreversibly, and azadirachtin itself competes for the site at 1.55 x 10^-8 M ⁶⁴. In fruit-fly cells it halved proliferation at 1.7 x 10^-7 M over 48 hours and damaged nuclei within six hours, and a binding complex fished out of those cells contained heat-shock protein 60, a mitochondrial chaperone; the authors called the complex 'putative' ⁶⁵. In developing locust sperm the label sat on the tails, and motility fell at 10^-4 M ⁶⁷. Against those findings stands the summary of the 2021 review, which, after listing effects on neuroendocrine activity, reproduction, feeding, apoptosis, protein synthesis and the genes for the ecdysone-synthesizing enzymes, concludes that 'the exact molecular mechanism of action of azadirachtin has yet to be fully understood' ¹⁵. Morgan said the same in 2009 ¹¹.

Mammals have none of this machinery, and the measurements agree. Cultured rat dorsal root ganglion neurons, the sensory cells of the spinal nerves, were 'relatively insensitive to azadirachtin compared with some invertebrate preparations': the changes in potassium currents and action potentials appeared only at 10 to 100 microM, a hundred to a hundred thousand times the insect-active range, and were reversible ²¹. Boar sperm, the one mammalian cell that swims with a flagellum, slowed above 10^-5 M ⁶⁷. The practical consequence is the subject of the safety section: rats fed technical material at up to 1.5 grams per kilogram a day for 90 days, about 180 milligrams of azadirachtin itself, showed nothing ¹⁹.

The same machinery that spares mammals does not spare other invertebrates, which is why neem keeps being proposed against parasitic worms. The dog heartworm Dirofilaria immitis molts in culture under the influence of 20-hydroxyecdysone, and azadirachtin at 10^-5 M prevented most third-stage larvae from reaching the fourth stage ¹¹⁴. That is a single in vitro experiment on a dog parasite, and no study on this page tested azadirachtin against a human intestinal worm in a person. In soil, azadirachtin reduced root-knot nematode reproduction in greenhouse tomato and nematode density in field cucumber in 2025 ⁴². In the malaria parasite it acts not through hormones but through the cytoskeleton: it left DNA replication and tubulin assembly intact but prevented the microtubules from organizing into mitotic spindles and the axonemes of male gametes, so the parasite could not make the motile cells it needs to infect a mosquito ⁶⁸. Lice are a different case altogether: the neem-seed shampoo that cures them does so by filling their breathing tubes, a physical suffocation that has nothing to do with azadirachtin ¹¹⁰.

The pesticide, the 1985 registration and the European patent

Azadirachtin is the most successful botanical insecticide to come out of the Meliaceae ¹⁵, and the slowest. The first commercial product was approved by the United States Environmental Protection Agency in 1985 for certain nonfood uses: Margosan-O, an ethanolic concentrate of neem seed extract standardized to 3,000 parts per million of azadirachtin, plus or minus 10%, made by a process developed at the Department of Agriculture in Beltsville, Maryland, and registered by Vikwood, Ltd. of Sheboygan, Wisconsin ²⁴. In 2009, twenty-four years after that registration, Morgan, who had helped find the molecule, was asking why it had 'still not achieved a prominent place among pesticides' and remained unlicensed in many countries, and set out to understand its failure to capture a larger market ¹¹. The European Union approved azadirachtin on 1 June 2011 for insecticide uses only, and in 2018 the European Food Safety Authority, peer-reviewing an application to extend the approval to mite control on greenhouse ornamentals, kept the acceptable daily intake of the extract at 0.1 mg per kilogram of body weight a day, named the liver as the target organ in the 90-day rat study behind that figure, and listed what is still missing: adequate long-term and carcinogenicity studies, a rabbit developmental study, data on how much of the extract is absorbed, and the fate of the eleven other limonoids and the non-limonoid components of the extract ¹⁸. In 2025 and 2026 it is still being tested on the margins: against root-knot nematodes in organic cucumber, where it was the one product that worked on a calendar schedule ⁴²; as a mosquito biopesticide that 'inhibits ecdysone signaling, reducing fecundity' in a 2026 review from India's malaria institute ¹¹⁵; and as a 1% spray that cleared mites from forty venom-farm scorpions without killing one, where the synthetic acaricides permethrin and fipronil are contraindicated because scorpion and mite are both arachnids and even minimal exposure can be fatal to the scorpion; the study had no untreated comparison group ¹¹¹.

Its environmental record is worse than its reputation. In a lacewing, a beneficial predator, neem oil eaten through poisoned prey at the larval stage left the adult midgut with lysed cells, swollen mitochondria and a stress response that did not fully repair the damage ¹¹⁶. Adult zebrafish exposed to a commercial neem formulation at 20 or 40 microliters per liter moved less and showed anxiety-like behavior, though learning was unaffected ¹¹⁷. A 2025 review of aquatic toxicity concludes that although azadirachtin degrades rapidly in soil, it has 'significant lethal and sublethal effects' on fish and aquatic invertebrates and that the fate of its metabolites and formulations is uncertain ¹¹⁸. In 2026, azadirachtin fed to juvenile Nile tilapia for 45 days at 20 to 170 mg per kilogram of feed damaged the structure of the kidney and the liver, raised a marker of oxidative damage in liver tissue and led its authors to advise caution about putting it in fish feed, even though it also raised the fish blood's activity against a fish pathogen ¹¹⁹. The 2021 in vivo toxicity review puts it plainly: for aquatic organisms, neem extracts were the least toxic and neem pesticides moderately to highly toxic, so commercial products 'should be handled with care' ⁶⁶.

The patent case is the part of neem's history that non-scientists know. A European patent, publication number 0436257, on a fungicidal preparation from neem seed was granted in 1994 or 1995, according to differing press accounts, and aid organizations and Indian groups announced their challenge to the neem patents in 1995 ²⁹; the European Patent Office revoked it on 11 May 2000 after a challenge brought by three organizations including a European organic-farming federation and an Indian environmental activist's group, and on 8 March 2005 the Technical Board of Appeal, in decision T 0416/01, dismissed the proprietor's appeal, finding the main request to lack an inventive step because it was obvious to try such formulations and the auxiliary request to add matter beyond the application as filed ²⁵. The proprietor by then was Thermo Trilogy Corporation; press reports describe the original grant as held jointly by the United States Department of Agriculture and the chemical company W. R. Grace, and the same corporate lineage, Vikwood to Grace to Thermo Trilogy, appears on the United States labels of Margosan-O ^24,25. Nature reported the first revocation in May 2000, under a headline crediting environmental campaigners with it ²⁸, and Nature Biotechnology reported in May 2005 that the appeal decision had revived the biopiracy debate ²⁶. The bioethics literature has treated the case ever since as the type specimen of a biopiracy allegation, and Hamilton's 2006 analysis argues that the activists' use of the word involved 'a deliberate simplification of science and IPR', that is, of intellectual property rights,, the point being less that an invention had been stolen than that the system for deciding what counts as invented was itself in question ²⁷. The ground of the decision was the ordinary patent-law test of obviousness, not the traditional knowledge of Indian farmers as such. The decision's own wording is known here only from a published summary of its text; its dates and outcome are corroborated by the contemporaneous reports in Nature and in Nature Biotechnology ^25,26,28.

The mouth: chewing sticks and the chlorhexidine trials

Cleaning the teeth with a twig is one of neem's oldest uses, and the mouth is where it has been tested most. The pharmacology is plausible. A water extract of neem chewing stick inhibited Streptococcus mutans, the bacterium most closely linked to tooth decay, or caries, and Streptococcus faecalis, which has been called Enterococcus faecalis since 1984, on blood agar, a nutrient jelly for growing bacteria, at 50% concentration, as did the Arabian arak stick ⁸¹; a 40% water-soluble azadirachtin preparation matched 0.2% chlorhexidine against S. mutans grown from the plaque of 30 children, both in zones of inhibition, the clear rings where bacteria fail to grow around a test substance, and in counts from a biofilm, which is a bacterial film of the kind plaque is ⁸². Both are agar results at concentrations no mouthwash reaches, and the one biofilm study that grew four oral species on extracted teeth for 21 days found 2% chlorhexidine significantly better than a neem extract made with methanol, an alcohol solvent ⁸³. A 2026 systematic review of 31 studies of traditional oral hygiene found miswak, the chewing stick of the arak tree, comparable to or better than toothbrushes for plaque and gingival inflammation; its neem evidence is not about neem sticks at all but about herbal toothpastes and rinses containing neem, clove, turmeric and ginger, which were as effective as fluoridated products in the short term, with 'methodological limitations and variability among studies' requiring 'cautious interpretation' ¹²⁰.

The mouthwash trials are the largest body of human evidence for neem and they are small. The table below lists the ones this page found with their sizes. The pattern is consistent: over three or four weeks, a 2% neem rinse reduces plaque and gingival indices about as much as 0.2% chlorhexidine, the standard antiseptic, in groups of 10 to 35 people ^{121,122,123,124}. Two trials cut the other way. In 80 adolescents randomized to four rinses for two weeks, chlorhexidine, an essential-oil rinse and povidone-iodine significantly reduced plaque, gingival scores and two gingival cytokines, which are signaling proteins of inflammation, and neem reduced them 'to a certain level' that was not statistically significant ¹⁰⁷. In the largest trial, 105 schoolchildren randomized to neem, mango or chlorhexidine for 21 days and followed for three months, all three reduced plaque and gingivitis, but only chlorhexidine kept the plaque reduction at two months and it had 'superior antigingivitis properties' ⁵⁶. The 2016 systematic review searched three databases and found, among 206 articles, three randomized trials that qualified; it could not pool them, and judged 'the quality of reporting and evidence along with methods of studies' to be 'generally flawed with unclear risk of bias' ³⁰. A 2022 review of herbal rinses in general reached the same verdict, 'low-quality trials' ¹²⁵. The newest trial, in 2026, randomized 44 men to a propolis-neem-chlorophyll rinse, a propolis-green-tea rinse, chlorhexidine or distilled water for seven days and found no difference between any of them, water included, on clinical indices or on calprotectin in gingival fluid ¹²⁶.

Commercial products complicate the reading. A six-month trial of 180 adults found a herbal toothpaste containing neem among six plant ingredients plus zinc and fluoride reduced plaque, gingival and bleeding scores by 23.5%, 25.6% and 73.3% more than a fluoride paste ¹²⁷; the companion study names the product and the comparator as both made by Colgate-Palmolive India ¹²⁸. Nothing in such a design can be attributed to neem, and the sponsor made both pastes. In dentistry's other corner, root-canal irrigation, neem has been tried against sodium hypochlorite, which is dilute bleach and the standard irrigant: in 50 Egyptian patients it lowered post-operative pain significantly at one of eight time points and reduced endotoxin, a toxic component of bacterial cell walls, by 18% against 8%, with neither irrigant clearing it ¹²⁹; in 80 Indian children it ranked last of three active irrigants behind nanosilver and hypochlorite for reducing anaerobic colony counts, the counts of bacteria that grow without oxygen ¹³⁰; and in a 2025 split-mouth trial of 15 children, in which each child's two sides of the mouth received different treatments, a sequence of miswak, the arak-tree chewing stick, then neem, tea tree oil and phytic acid matched the conventional sequence ¹³¹. A 2023 systematic review of ten clinical studies in about 442 patients concluded that the natural irrigants cannot replace hypochlorite routinely, although neem was among those that came out ahead of it in individual comparisons and was associated with less pain afterward ¹³². Used instead as a dressing left inside the canal between visits, neem 'showed poor results' against calcium hydroxide and Triphala in a 2025 randomized trial of 240 patients ¹³³, and chilled neem irrigant matched chilled hypochlorite for post-operative pain in a 2026 trial of 108 patients ¹³⁴. Cavities have been counted once: in 120 children aged three to five, a toothpaste containing neem, clove and 500 parts per million of fluoride produced a smaller mean increase over twelve months in decayed, missing or filled baby teeth than a plain 500 parts per million fluoride paste, 0.42 against 0.95, with new lesions in 16.6% against 28.3%; that trial cannot separate neem from clove and appeared in a conference-style journal supplement ¹³⁵. The conclusion the evidence supports is narrow: a neem rinse is a reasonable short-term adjunct to brushing for someone who cannot tolerate chlorhexidine's taste and staining, in the knowledge that it is probably somewhat weaker, that neem alone has never been tested against caries, and that no trial has measured tooth loss or anything that happens after three months ^30,56,135.

Skin and the parasites that live on it: scabies, lice, sand fleas, psoriasis, repellents

Scabies had no randomized trial until 2025. In 1992 a neem-turmeric paste was used on 814 people in Tamil Nadu with 97% cured in 3 to 15 days and no adverse reaction noticed; there was no control group and the authors asked for further research ¹³⁷. The mite is killed by the oil in a dish: an extract of neem oil made with petroleum ether, a light solvent, had a median lethal concentration of 70.9 ml/l at 24 hours against rabbit scabies mites, and two of its four chromatographic fractions, the portions separated from it by how they travel through a column, were active ¹³⁸. The randomized trial arrived in December 2025: 61 male students at Indonesian boarding schools, where scabies recurs, were randomized to 5% permethrin, 10% neem leaf lotion or 10% lotion of brotowali, an Indonesian medicinal vine; lesion counts fell in all three with no difference in the physician's global assessment, quality of life improved earlier with the herbal lotions, and no adverse effects were reported; the authors state that adherence and application time 'could not be fully controlled' and reinfestation 'could not be excluded' ⁹⁹. About twenty per arm, in a setting where permethrin itself may be failing, which is both the trial's reason and its limit.

Head lice are where neem's best-run human trial took place, and the mechanism there is not the molecule's. A neem-seed shampoo killed all stages of lice on 60 heavily infested Egyptian children in 10 to 30 minutes without irritation ¹³⁹; five minutes' immersion stopped every louse egg from hatching, which the authors attributed to blocking of the eggs' air pores ¹⁴⁰; and microscopy showed the product filling the louse's tracheal system, so that death came by suffocation within 3 to 10 minutes ¹¹⁰. The randomized, investigator-blinded trial of 2017 in 119 Egyptian children gave cure after a single application in 60 of 60 (100%, 95% confidence interval 94.04 to 100) with the neem shampoo against 54 of 57 (94.74%, 85.38 to 98.90) with dimethicone, p = 0.112, and a combined success rate of 100% versus 90.7%, p = 0.024 ¹⁰⁹. A 95% confidence interval is the range within which the true cure rate probably lies, and p is the probability of seeing a difference at least this large by chance alone. The trial's stated aim was to show a cure rate of at least 85% after one application, which both products met; it was not designed to show that neem beat dimethicone ¹⁰⁹. Two cautions: dimethicone is itself a physical pediculicide, so the comparison is between two suffocants, and an employee of the manufacturer, Alpha-Biocare, is an author of this trial and of an earlier uncontrolled series of 65 patients treated with the same product ^109,141. The product's virtue, that it needs no insecticide, is also the reason it tells us nothing about azadirachtin.

Tungiasis, the sand flea that burrows into the feet of children in coastal Kenya, got a registered randomized trial in 2019: 96 children treated with 20% neem seed oil in coconut oil or with potassium permanganate foot baths. Neem killed about 40% of embedded fleas by day six, no better than permanganate, but the fleas aged faster (odds ratio 3.4, meaning the odds of that outcome were 3.4 times higher with neem; 95% confidence interval 1.22 to 9.49), local pathology fell and children were more often pain free (odds ratio 3.5, p = 0.001) ¹⁴². A proof of principle on secondary outcomes, as its title says.

Psoriasis rests on reviews; the trials underneath them are not in PubMed. A 2017 review of twelve controlled and uncontrolled trials of oral botanicals found the evidence 'limited in quantity and quality' and neem extract among the three agents that seemed most efficacious ¹⁴³; a 2015 dermatology review put neem among herbals with 'a smaller amount of evidence' than Mahonia and indigo, and warned of systemic toxicity and contact dermatitis from herbal supplements ¹⁴⁴. The underlying study, described in a 2016 review, was a double-blind trial of an aqueous leaf extract added to coal tar in 50 patients, with a quicker response than placebo ¹; it is not indexed in PubMed. A neem-turmeric-Cassia soap from a Unani institute cleared ringworm on microscopy in 80% versus 20% of 30 patients over four weeks in a single-blind trial ⁵⁹, and in 2025 a neem-propylene glycol-salicylic acid lotion improved arsenical keratosis of the palms, the hard bumps caused by years of arsenic in drinking water, in a 30-patient comparison in Bangladesh ¹⁴⁵; both are small and neither isolates neem. The skin's clearest signal is the one against the oil: allergic contact dermatitis, an itchy rash from an immune reaction to something touching the skin, reported in 2016 in a psoriasis patient sensitized to three herbal oils ⁴⁷, in a Dutch man whose physicians noted that only three such patients had been published before him ⁴⁸, airborne in a Spanish man in a letter published in 2018 ⁴³, in Italian and Japanese patients in 2019 and 2020 ^49,50, occupationally in a cosmetics worker in 2020 ⁴⁴, in an eczema patient who used it on a flare in 2023 ⁴⁵, and in a 2025 Belgian case whose authors call it 'an emerging cause' ⁴⁶. Chewing the leaves has caused the same reaction inside the mouth ⁵¹; self-treatment with the oil was reported in 2023 to have worsened a bullous pemphigoid, an autoimmune blistering disease ¹⁴⁶; and six months of rubbing it on the back left a 77-year-old man with a pale rash ¹⁴⁷.

Against malaria mosquitoes, the repellent claim is the one most plainly reversed. In 1993 India's Malaria Research Centre reported that 2% neem oil in coconut oil on the skin gave 'complete protection for 12 h' from all anopheline mosquitoes, meaning the Anopheles species that carry malaria ¹⁰³. In 2002, in human landing catches against Anopheles darlingi in the Bolivian Amazon, a eucalyptus-based repellent gave 96.89% protection for four hours, deet, the standard synthetic repellent, gave 84.81%, and the neem-based product 'did not provide significant protection' ¹⁰². In 2008, on rabbit skin against Anopheles stephensi, the dose for 50% repellency was 0.007 mg/cm2 for permethrin, 0.005 for deet and 0.191 for neem, 'the least effective agent' ¹⁴⁸. The 1993 note stands as a claim; the two later studies with comparators are the measurement. Against sand flies, which carry leishmaniasis, the result went the other way: in Ethiopia, 2% and 5% neem oil in coconut oil gave 96.28% protection for 7 hours 20 minutes and 98.26% for 9 hours against Phlebotomus orientalis in the laboratory, with similar protection in the field ¹⁴⁹. A 2026 systematic review of 19 human studies published between 2015 and 2025 puts the plant-based repellents, neem among them, at two to four hours of complete protection against five to nine hours for deet ¹⁵⁰. The honest summary is that neem oil repels some biting flies for a few hours and does not hold off malaria mosquitoes the way the 1993 note claimed.

Malaria: a dish, a mouse, and no patient

Neem is one of the plants people across West Africa and South Asia take for fever, and the laboratory has given the practice some support. In 1997, of 22 Meliaceae species screened against two clones of Plasmodium falciparum, twelve were active, including neem leaf, whose active principle was gedunin, the only compound in the study with better activity than chloroquine, the old standard antimalarial drug, against the clone the paper calls W2 and labels chloroquine-sensitive, although W2 is the standard chloroquine-resistant laboratory line ³⁹. Fractions of a Nigerian leaf extract inhibited the asexual parasites, the blood stage that causes the fever, at concentrations as low as 10^-5 microg/ml and lysed half of developing gametocytes at 10^-3 microg/ml, with the authors' own proviso, 'if they are found safe and effective in vivo' ⁴⁰. Among 13 Kenyan antimalarial plants, a methanol extract of neem was one of the three most active in vitro, halving parasite growth below 10 micrograms per milliliter, an IC50 under that figure, against both the chloroquine-sensitive and the chloroquine-resistant clone, and gave 89% parasite suppression in mice ³⁸. In 2026 a Ghanaian group found a water extract of neem synergistic with Cryptolepis sanguinolenta, a West African root long used against fever, against both strains in a dish ¹⁵¹. A 2022 review of 124 Meliaceae compounds found limonoids the most active and gedunin 'equal to quinine', and asked for the absorption, distribution, metabolism, excretion and toxicity studies a drug would need ¹⁵². The largest count of the African work, a 2021 systematic review of 722 experiments on 502 plant species drawn from 200 papers, found Azadirachta indica the most frequently investigated species of all, and also the species most frequently reported toxic and among those most frequently reported inactive ¹⁵³.

The most interesting animal result is about transmission, not cure. Azadirachtin leaves the asexual blood stages alone but stops the male gametocyte from making motile gametes, because the microtubules that should form spindles and flagella cannot organize ⁶⁸. In 2010 a standardized seed extract, NeemAzal, injected into the abdominal cavity of gametocyte-carrying mice at 50 mg/kg azadirachtin one hour before mosquitoes fed, blocked Plasmodium berghei completely: no oocysts in any of 138 mosquitoes across three experiments, and none of six healthy mice bitten by those mosquitoes became infected, while the midgut smears showed fewer zygotes and no mature ookinetes ⁶¹. Half that dose did nothing, and so did a commercial 'azadirachtin' at the same nominal dose, until chromatography showed it held about half the azadirachtin claimed ⁶¹. In mice with cerebral malaria, an ethanolic extract at 300 to 1,000 mg/kg reduced brain swelling and neuronal death comparably to chloroquine and to artemether, a drug derived from artemisinin, by histology rather than survival ¹⁵⁴.

There is no clinical trial. The 2025 review of natural transmission-blocking agents names Azadirachta indica among extracts with 'potent gametocytocidal effects' and states that 'despite promising in vitro and in vivo data, the transition of these compounds to clinical applications remains limited' ⁹⁵; the 2026 count of registered neem trials found none in malaria or any systemic disease with a phase II/III result ⁸. The authors of the 2010 mouse study explained why in their discussion: limonoids have a short half-life and poor bioavailability, limonin was undetectable in human plasma 24 hours after a 30 mg/kg dose of its glucoside, and gedunin's failure to cure in vivo had been ascribed to poor absorption from the gut ⁶¹. A leaf extract that works in a dish has to reach the blood to work in a person, and the first test of that failed: nimbolide and crude neem extracts that inhibited the parasite in culture did nothing to infected mice, by mouth or by injection ⁹². Nothing on this page shows that any neem preparation reaches an active concentration in human blood.

Diabetes, ulcers, COVID-19: the systemic claims at true size

Diabetes is the claim with the longest pedigree and the thinnest human base. Neem is among the twelve most-cited antidiabetic plants of West Africa and sits in the Ayurvedic record of remedies for diabetes ^5,41. In rodents made diabetic with streptozotocin, extracts lower blood glucose, as the 2026 review summarizes and a 2026 mouse study again shows ^8,155. In pregnant rats, neither neem seed oil nor azadirachtin had any hypoglycemic or antihyperglycemic effect, in diabetic or non-diabetic dams, and both impaired fetal growth ¹⁰⁸. The one randomized human trial is from 2021: 80 people with metabolic syndrome, 20 per group, took placebo or 125, 250 or 500 mg of an aqueous leaf-and-twig extract twice daily for 12 weeks, and fasting and post-meal glucose, insulin resistance (the body's weakened response to its own insulin), HbA1c, endothelial function (how well the lining of the blood vessels relaxes) and inflammatory markers improved in a dose-dependent way ⁹⁸. It is registered, placebo-controlled and single-center, the abstract reports no absolute numbers, and it has not been repeated. The 2026 West African review's sentence covers it: 'few of these plants have undergone clinical trials and in-depth toxicological evaluations' ⁴¹.

Peptic ulcer has one uncontrolled human study, from Kolkata in 2004: patients with acid-related complaints given 30 mg of freeze-dried, or lyophilized, bark extract twice daily for 10 days had 77% less gastric acid secretion, and at 30 to 60 mg twice daily for 10 weeks duodenal ulcers 'almost completely healed' on barium meal, an X-ray of the stomach and duodenum taken after swallowing barium, or on endoscopy, with one esophageal and one gastric ulcer also healed and blood tests normal ¹⁵⁶. The numbers of patients are not in the abstract, there was no control arm, and in the era of Helicobacter pylori eradication, which cures most ulcers by killing the stomach bacterium that causes them, and of the proton-pump inhibitors, the drugs that shut off stomach acid, nobody has followed it up. The liver has mouse data in both directions: a 2025 ethanolic leaf extract containing azadirachtin and nimbolide reduced mortality and transaminases, the liver enzymes that leak into the blood when liver cells are damaged, in a mouse model of acute hepatitis ¹⁵⁷, and a 2016 review of herbal hepatotoxicity lists Azadirachta indica among the plants 'found to be hepatotoxic', meaning damaging to the liver, which the oil's record in children explains ¹⁵⁸.

COVID-19 produced the newest claim. A Cochrane review, a systematic review produced by the international Cochrane collaboration, found in September 2020 no completed study of any antimicrobial mouthwash or nasal spray in the disease, and listed among its 16 ongoing studies one testing a spray containing neem oil and St John's wort and one testing a neem extract ¹⁵⁹. The one published result is a pilot: 190 health workers and relatives of patients randomized to 50 mg of a 'proprietary, patent-pending' leaf extract or placebo twice daily for 28 days; 13 of the 190 tested positive in all, and among the 154 who completed the study as planned the split was 3 on neem against 8 on placebo, all of them without symptoms, a relative risk of 0.45 ⁹⁷. Eleven events, a per-protocol analysis, a product with a patent pending and a conclusion of 'potential': a hypothesis, which its authors say warrants further trials and which none has since tested. A 2026 review from an Indian institute lists 'only fifteen interventional trials' registered for neem in all, most for the mouth, 'with no published phase II/III results for systemic diseases' ⁸.

The contraceptive and microbicide program: animals that worked, women who were not tested

The longest-running human program on neem was G. P. Talwar's, at the National Institute of Immunology in New Delhi, and it is instructive because it stopped exactly where efficacy would have been measured. The animal results were striking. A single 100-microliter dose of neem oil placed in the uterus of a fertile rat blocked fertility for 107 to 180 days, the block was confined to the treated horn when only one was treated, half the animals regained fertility and delivered normal litters by five months, and the ovaries kept cycling; the mechanism was a pre-implantation block, meaning the embryo failed before it could attach to the womb, with leukocytes infiltrating the uterine lining ⁸⁶. One milliliter in a bonnet monkey's uterus did the same for 7 to 12 months, reversibly, with more antigen-presenting cells, the immune cells that display foreign material to other immune cells, in the endometrium ⁸⁷; the embryos degenerated on the treated side between days 3 and 5 after mating ¹⁶⁰. Oral Praneem, a purified seed extract, given for six days from day 25 of the cycle terminated established early pregnancies in baboons and bonnet monkeys, with falling chorionic gonadotropin and progesterone, the hormones that sustain an early pregnancy, while peanut oil controls stayed pregnant ¹⁰¹. On human sperm in a test tube, the purified extract immobilized every sperm within 20 seconds at a 25% concentration, and was synergistic with saponins of the soapnut and with quinine, the three ingredients of Praneem cream ^100,161. When the group fractionated the seed to find what blocked fertility, the answer was a mixture of six free fatty acids and their methyl esters, not azadirachtin or any limonoid ⁸⁸.

The human studies were about safety. In 1995 purified neem oil was instilled once into the uteri of 18 women who had already been sterilized by tubal ligation: no untoward reaction, menstrual pattern and ovulation unchanged, one symptomless endometritis, an inflammation of the womb lining, on biopsy, a transient eosinophilia, a rise in one class of white blood cell, in two, and an unimpaired antibody response to an experimental birth-control vaccine given alongside ⁸⁹. Sterilized women cannot show whether a contraceptive works, and no fertility trial followed in PubMed. The vaginal tablet, repositioned as a candidate microbicide against HIV, went through a phase I study in Pune in which nine of 20 women (45%) had 17 episodes of genital irritation ⁹⁰, and a phase II trial in which 100 low-risk women were randomized, 50 to the tablet and 50 to placebo, and used the assigned product before each act of sex for six months, with only transient discomfort, no serious adverse event and no HIV seroconversion, meaning that no participant became infected, and whose authors, writing after nonoxynol-9, cellulose sulfate and Carraguard, three earlier vaginal products, had all failed to prevent HIV in trials, concluded that 'additional preclinical evaluations should be undertaken urgently before moving into effectiveness studies' ⁹¹. Of the 95 women who completed all follow-up visits, 67, or 70.5%, reported using the product at every act of sex; odor was a concern in both arms, and women on Praneem were more likely to be only partly adherent because of how it felt during sex (adjusted odds ratio 3.8, meaning odds 3.8 times higher after accounting for other differences; 95% confidence interval 1.4 to 10.8) ¹⁶². No effectiveness study was run.

The same tablet was tried for the syndrome of abnormal vaginal discharge, and the three trials show how the design shapes the answer. In an uncontrolled multicenter study of 141 women, 97% reported relief and 74% were judged cured on speculum examination, with microscopic cure in every case of Trichomonas, a sexually transmitted parasite, in 77% of cases of Candida, a yeast, and in 68% of cases of bacterial vaginosis, an overgrowth of the wrong vaginal bacteria, at the cost of transient burning in 55% ¹⁰⁴. Against a povidone-iodine pessary, a tablet inserted into the vagina, in 99 women, symptom relief was 92% versus 81.6% ¹⁰⁵. In an independent randomized trial of 100 women in Chandigarh against a pessary of clotrimazole, tinidazole and lactobacilli, symptomatic relief was 78% versus 86%, only 36% of the women had any laboratory-confirmed infection, neither product was efficacious in candidiasis, the drug pessary cured all cases of bacterial vaginosis while Praneem showed a non-significant trend, and vaginal irritation occurred in 16% on Praneem against 4% ¹⁰⁶. Praneem belongs with the other unfinished microbicides of its era, and the lesson of the whole program is that the active part of the seed is the oily part: the intrauterine antifertility fraction was a mixture of free fatty acids rather than any limonoid ⁸⁸, an oral purified extract ended established pregnancies in two primate species, though which fraction did it was not identified ¹⁰¹, and fatty acids are one of the proposed causes of the infant poisonings ^72,88. There are no human data on neem and pregnancy, but a preparation that aborts primates is reason enough for a pregnant woman not to swallow neem seed products ^69,101.

• Arterial blood gas, glucose, ammonia and liver enzymes in a child who has had neem oilNeem oil poisoning is a clinical diagnosis from the history; the laboratory confirms the pattern. The founding series of thirteen Malaysian infants recorded vomiting, drowsiness, metabolic acidosis, neutrophil leukocytosis and encephalopathy within hours ²²; the rat model adds high ammonia and transaminases and low glucose ⁷³; the one adult case report in full gives pH 7.34, bicarbonate 16 and a normal lactate ¹⁰.In any infant or child with vomiting, drowsiness and seizures where a tonic, oil rub or worm remedy has been givenAcidosis and leukocytosis are the consistent laboratory findings across the reported series ^22,33; glucose was low in rats, and in the one treatment study, done in mice, glucose improved survival significantly, carnitine slightly, and the two together best, while steroids did nothing ^73,85Nothing in the blood is specific; aflatoxin poisoning and Reye's syndrome itself had to be excluded in the fatal 1982 case ³⁶, and there is no assay for the oil or azadirachtin in clinical use
• Liver biopsy or autopsy with electron microscopyFatty infiltration of liver and proximal kidney tubules with mitochondrial damage was the finding in the one biopsied infant and in the poisoned mice ²², and autopsy of a fatal newborn case gave the postmortem picture ³⁶.Historically, to establish that the illness was Reye-like; today only at autopsyMicrovesicular steatosis with swollen mitochondria is the lesion, and the mouse time course shows it forming within an hour ⁷²It cannot distinguish margosa oil from the other mitochondrial poisons that open the same permeability pore ⁷¹
• Brain magnetic resonance imaging (MRI)In an adult with visual loss after neem oil, MRI showed symmetrical signal change in both putamina extending into the posterior limb of the internal capsule; her laboratory tests were normal and she recovered completely, though no follow-up imaging is reported ^10,54. In a second woman who drank about 150 ml, the discs were swollen, the same putaminal change was present, and vision came back only to 20/200 at two months ⁵⁵.In encephalopathy that does not clear, or with focal signsOne case report; the pattern is that of a metabolic basal-ganglia injury rather than anything specificThe 73-year-old man's MRI showed only old small-vessel change despite seizures and four days of drowsiness ¹⁰
• High-performance liquid chromatography for azadirachtin contentNeem products are standardized, when they are, by high-performance liquid chromatography (HPLC), a laboratory method that separates a mixture and measures each component, for azadirachtin A; the transmission-blocking study found by this method that a commercial 'azadirachtin' contained about half the amount declared, which explained why it failed where the standardized extract worked ⁶¹. The contraceptive fraction was likewise tracked by an analytical HPLC method ⁸⁸.Quality control of pesticides, research extracts and, rarely, medicinal productsGood for azadirachtin A; it says nothing about the fatty acids that carry the toxicity of the oil or the antifertility effect ^10,88Almost no clinical product used in the trials on this page reports an assayed content, which is one reason the trials cannot be compared ³⁰
• Plaque and gingival indices, mite counts and louse combingThe human efficacy evidence rests on these: the Loe-Silness gingival index and plaque index in the mouthwash trials ^56,121, lesion counts and the physician global assessment in the 2025 scabies trial ⁹⁹, and combing for live lice at 1-2 hours and days 5 and 13 in the louse trial ¹⁰⁹.As outcome measures in the only randomized trials of neem in peopleReproducible in the hands of trained examiners; the louse trial reported cure as 60/60 with a 95% confidence interval ¹⁰⁹A plaque index at 21 days does not measure caries or tooth loss; the only trial that counted cavities tested neem combined with clove and fluoride, so neem by itself has never been tested against either ^30,135

Margosa oil poisoning: the whole record, infant to adult

The 1981 Lancet report is the beginning of the written record. Thirteen infants in Malaysia, of Indian families, developed vomiting, drowsiness, metabolic acidosis, a rise in neutrophils and encephalopathy within hours of ingesting margosa oil; a liver biopsy in one and the necropsy, or autopsy, of mice given the oil showed pronounced fatty infiltration of the liver and of the proximal tubules, the first filtering segments of the kidney, swelling of the brain, which physicians call cerebral edema, and mitochondrial damage on electron microscopy, and the authors proposed that the oil was a cause of Reye's syndrome among Indians in Malaysia ²². A fatal newborn case with postmortem findings followed in 1982, in which aflatoxin poisoning, from a liver-damaging mold toxin found in spoiled grain and nuts, had to be excluded as the alternative explanation ³⁶. The Indian series of the same year is summarized by later authors as 12 children given single doses of 25 to 60 ml, of whom 10 died ^10,23. Singapore reported two infants in 1990 with the summary that stands as the clinical description: vomiting, drowsiness, rapid breathing and recurrent generalized seizures, leukocytosis and metabolic acidosis in the laboratory, management aimed at the convulsions, prognosis usually good 'but fatalities and neurological deficits have been reported', and the oil 'often administered orally to neonates and infants regularly in small amounts' ³³. By Mishra and Dave's reckoning the Singapore group had seen 22 infants given a few drops to 5 ml, all of whom recovered ¹⁰.

Why the oil does this was worked out in rats, mice and test tubes over the following decade, mostly by Sinniah's group with American and Japanese co-authors, and in 1996 by an independent mitochondrial laboratory ^{70,71,73,85,163}. Rats injected twice with the oil developed florid neurological signs with high ammonia and transaminases and low glucose, a fatty liver with glycogen depletion and abnormal mitochondria, but, unlike Reye's syndrome, their hepatic mitochondrial enzymes did not fall and their brains did not swell ⁷³; a separate rat study did find raised brain water and seizures in every animal ¹⁶³. On isolated rat liver mitochondria the oil behaved as an uncoupler: resting respiration rose, the respiratory control ratio, which measures how tightly fuel-burning is coupled to energy capture, fell along with the content of ATP, the molecule cells use to carry energy, the balance between charged and spent electron carriers in the respiratory chain, its redox state, changed, and the proportions of coenzyme A esters shifted toward those of disturbed fatty-acid metabolism; coenzyme Q and carnitine partly relieved the block ⁷⁰. A time course in mice put the first changes in liver-cell nuclei at 30 minutes and in mitochondria at 60 minutes, followed by glycogen loss, proliferation of the endoplasmic reticulum, the cell's membrane factory, and microvesicular fat, with fatty acids mobilized from body fat and fat globules from the gut adding to the load; the authors describe the oil as 'a long-chain fatty acid compound' ⁷². In 1996 neem oil was shown to open the cyclosporin-sensitive permeability pore of liver mitochondria, the lesion shared by salicylate, valproate and the other chemicals implicated in Reye's syndrome and in Jamaican vomiting sickness, the poisoning caused by unripe ackee fruit ⁷¹. Treatment was tested in the mouse model because, as the authors wrote, controlled trials in children were not feasible: glucose improved survival significantly, carnitine slightly, their combination best; coenzyme Q10 did not alter the liver pathology; and the steroids dexamethasone and methylprednisolone did nothing ⁸⁵. That is still the basis of management, which in every human report is supportive: control the seizures, give glucose, correct the acidosis ^10,33.

The Indian reports of the 2000s and 2010s add the long-term harm and the adult. A previously normal five-year-old in Pune developed refractory seizures and metabolic acidosis after neem oil, and the article's own subject index also lists a cardiac arrest; he was left with auditory and visual disturbance and unsteadiness of gait ⁵². A Sri Lankan infant was given the oil as a home remedy for intestinal worms and developed toxic encephalopathy ³⁴. An infant in Chandigarh was reported in 2014, the most recent pediatric case this page could find in PubMed ³⁵. Among adults, a 35-year-old woman who drank the oil in self-harm lost vision in both eyes and had symmetrical signal change in both putamina extending into the posterior limb of the internal capsule on magnetic resonance imaging, or MRI, with normal laboratory tests, and recovered completely, no follow-up imaging being reported ^10,54; another 35-year-old woman, reported in 2014 after drinking about 150 ml, had swollen optic discs, no perception of light in either eye and the same putaminal change, and her vision came back only to 20/200 at two months ⁵⁵; a 35-year-old woman who drank a neem-kernel pesticide in Vellore needed ventilation for neurotoxicity and survived without lasting harm, the authors noting that adult overdose of a product 'considered natural and safe' had not been described ⁵³; a man presented with severe metabolic acidosis, seizures, kidney failure and encephalopathy after a small dose and his physicians titled the report 'Herbal remedy is natural and safe--truth or myth?' ³¹; and a 73-year-old man with type 2 diabetes and psoriasis who swallowed 20 ml by accident vomited within the hour, convulsed in the emergency department, had a metabolic acidosis with normal lactate, liver and kidney function, stayed drowsy four days and went home in a week ¹⁰. Neem leaf poisoning in adults has a one-page letter from 1985 ¹⁶⁴ and a 2014 report from Mayotte of adults who developed neurological signs and multiorgan failure after traditional remedies containing neem or the related chinaberry, all of whom recovered and in none of whom the plant could be identified with certainty ¹⁶⁵; in chicks, feed containing 2% or 5% neem leaf for four weeks damaged the liver and kidney, and the damage had not fully healed two weeks after the leaf was withdrawn ¹⁶⁶. Cats have been poisoned by neem-oil flea products ¹⁶⁷.

What is missing from the record matters as much as what is in it. PubMed holds no report of a poisoned child after 2014, but the adult reports have gone on. In 2017 a young man in Delhi drank 100 ml of a household rat poison made of 40% azadirachtin oil with tea oil, pine oil and kerosene, and reached a methemoglobin level of 80%, meaning that four fifths of his hemoglobin could not carry oxygen; methylene blue and intravenous vitamin C brought it down and he went home after five days, though with kerosene and two other oils in the product neem cannot be singled out ¹⁶⁸. In 2025 a young woman with suspected neem-powder self-poisoning was made to vomit with salt water at home and arrived with a serum sodium of 187 mmol/L, a dangerous level, from which she recovered; the harm came from the remedy rather than from the poison ³⁷. The 2023 to 2026 literature adds two contact-dermatitis reports ^45,46, a bullous pemphigoid that worsened after self-treatment with the oil ¹⁴⁶, a pale rash after six months of rubbing it on the back ¹⁴⁷, and a study of 50 men working in neem-oil extraction in Tamil Nadu who had more cough, wheeze and nasal irritation and worse lung function than 50 unexposed men ¹⁶⁹. Nobody should read the gap in pediatric reports as the end of the poisonings. The oil is still given to infants, the case series come from a handful of teaching hospitals, and the 2026 review of neem's pharmacology treats 'pediatric encephalopathy from neem oil ingestion and embryotoxicity' as established and calls for 'explicit safety guidelines' ⁸. The dose that harms is small in absolute terms: a few drops to 5 ml in the Singapore infants who recovered, 25 to 60 ml in the Indian children who mostly died, 20 ml in a 73-year-old who was intubated ¹⁰. Which part of the oil does the harm is still unsettled. The physicians who reported the adult cases blamed azadirachtin ^10,53; but the pure molecule is nearly inert in rodents at these doses ^19,20, the oil and not the molecule malformed rat fetuses ⁶⁹, and the oil is a mitochondrial uncoupler of the same class as the metabolite of aspirin ^70,71. The weight of the animal evidence points to the oil's fatty fraction, which has been proposed as the cause but not proved ^72,88.

The safety of azadirachtin for mammals, as measured

The molecule has been tested the way a pesticide must be, and it passes. Technical azadirachtin of 12% purity given by mouth to male and female rats at 500, 1,000 and 1,500 mg/kg per day for 90 days produced no sign of toxicity, no death, no change in organ weights, no pathology and no change in serum or blood values, and the authors proposed the top dose as the base for a no-observed-effect level ¹⁹. Since the material was 12% azadirachtin, the pure-molecule dose at the top was about 180 mg/kg/day, which is still enormous. A two-generation study fed female rats 100 to 1,000 ppm of technical azadirachtin in the diet, about 5 to 50 mg/kg, and found no toxic effect in the parents, no teratogenic effect in two generations of offspring and no cumulative effect on development or reproduction ⁶⁶. The 2004 safety review from Wageningen, which gathered oral toxicity data in humans and animals for every kind of neem preparation, estimated a safe dose of 15 mg/kg per day for pure azadirachtin, against 0.002 and 12.5 microg/kg per day for the two non-aqueous extracts it assessed, which it found the most toxic class, 0.26 mg/kg per day for seed oil, and 0.3 mg/kg per day, or 2 microliters per kilogram, for aqueous extracts; its summary was that reversible effects on reproduction in both sexes were the most important toxic effects of sub-acute and chronic exposure, and that neem-derived pesticides, applied with care, 'should not be discouraged' ²⁰. A standardized commercial extract given to mice by mouth at 3.8 to 12.7 g/kg killed none in 14 days, though behavior changed above 8 g/kg ¹⁷⁰. European regulators set their numbers on the whole extract rather than on the molecule: an acceptable daily intake and an acceptable operator exposure level of 0.1 mg per kilogram of body weight a day, both derived from a 90-day rat study whose no-observed-adverse-effect level was 32 mg/kg per day with the liver as target organ, divided by a safety factor of 300 because adequate long-term, carcinogenicity and rabbit developmental studies are missing, and an acute reference dose of 0.75 mg/kg from a rat developmental study with a maternal no-effect level of 225 mg/kg per day; the three commercial extracts damaged chromosomes in cultured human lymphocytes but not in animals, and fertility and reproductive performance were unimpaired in a valid multigeneration rat study ¹⁸.

The reproductive findings are the ones worth knowing, and they distinguish the oil from the molecule. Rats given neem seed oil throughout pregnancy had more malformed or variant fetuses, especially visceral ones; rats given azadirachtin did not ⁶⁹. But both oil and azadirachtin raised the share of fetuses small for gestational age and raised lipid peroxidation in the mothers, and neither lowered glucose ¹⁰⁸. Azadirachtin-A injected under the skin of male rats at 1.5 mg/kg for 24 days reduced reproductive-organ weights and sperm quality, as did nimbolide at the same doses, while a neem-based insecticide at up to 45 mg/kg azadirachtin given to pregnant mice produced no malformations ⁶⁶. Boar sperm slowed in vitro above 10^-5 M ⁶⁷. In a fish, injected azadirachtin produced chromosome breaks and fusions, and the authors suggested the product 'should be viewed with caution' ⁶⁶. The 2021 in vivo review concludes that neem extracts have no or very low acute toxicity in mammals by mouth, that sub-acute and sub-chronic toxicity is dose-dependent and reversible, with kidney, liver and testis as the target organs of the oil at 1,600 mg/kg/day in mice and 177 mg/kg/day as the no-observed-adverse-effect level, and that most studies of reproduction 'described changes in both the male and female reproductive systems' ⁶⁶. One of that review's authors is affiliated with the World Neem Organization, which the paper discloses.

The human data on the molecule itself are a single case. The 35-year-old who drank a neem-kernel pesticide in self-harm had neurotoxicity severe enough to need ventilation and recovered completely ⁵³; the formulation and the dose of azadirachtin in it are not known, and the oil and solvents in a pesticide concentrate are the likelier culprits, given that pure azadirachtin at gram doses does nothing to a rat ¹⁹. There is no human pharmacokinetic study of azadirachtin, nor of nimbolide, gedunin or nimbidin ^61,80. What exists for the human body is therefore an inference: a molecule that acts through hormones and taste receptors humans lack, that is inert in rodents at doses far above anything a person could take in from treated food, and that disturbs sperm and dividing cells only at concentrations no diet or pesticide exposure reaches ^{19,20,21,65,67}. One claim about the molecule has been tested and rejected: a popular explanation for cannabinoid hyperemesis syndrome, the cycles of severe vomiting seen in some heavy cannabis users, blames neem oil or azadirachtin sprayed on the plants, and a 2024 review examined that explanation and refuted it, tying the syndrome instead to heavy use of potent cannabis and to inherited variation in the user ¹⁷¹.

The antimicrobial claims in vitro, and the shape of the hype

Neem extracts kill microbes on agar; almost everything does at the concentrations used. The question is how they compare, and when a comparator is present neem is seldom first. Against nine urinary pathogens resistant to several classes of antibiotics, methanolic neem leaf extract fell into the group of eight plants with 'moderate control capacity', behind ten other Indian timber trees ¹⁷². On a multispecies dental biofilm, 2% chlorhexidine was significantly better than neem ⁸³. In poultry feed in 2025, clove and cinnamon oils were strongest against Aspergillus and Fusarium and 'neem extracts exhibited moderate efficacy' ¹⁷³. Against the maize mold Fusarium verticillioides, lemongrass and basil oils stopped growth completely, while 'the oil of neem seeds showed no inhibitory effect but rather accelerated the growth' ¹⁷⁴. Against Aspergillus carbonarius, a mold that makes ochratoxin, a toxin that damages the kidney, neem oil inhibited growth 97.6% at 0.3% but only 40.2% at 0.5% and 64.7% at 1.0%, a dose-response that runs the wrong way above the lowest effective concentration, though toxin production was blocked at 0.1 to 0.3% ¹⁷⁵. The honest summary is that neem seed extracts have antifungal activity that depends on the solvent, the strain and the dose, and that the 2026 seed-extract study finding 45% reduction of fungal mass at 31.25 mg/ml and over 90% suppression of aflatoxin at 250 mg/ml is working at roughly 3% and 25% extract by weight, concentrations no tissue in a living body could reach ¹⁷⁶.

The literature's tone is part of the subject. A 2025 review is titled 'A Miracle Herb; Panacea for All Ailments' and calls the tree 'a natural remedy for numerous illnesses' ⁶²; a 2016 review that summarizes the infant poisonings concludes three sections later that the tree acts 'without any adverse effect' ¹; a 2004 trial report in a dental journal concluded from 48 open-label subjects that 'better therapeutic efficacy can be achieved using gels' ⁹⁶. Against that tone stand the papers that counted. Three qualifying randomized trials in 206 articles on mouthwash, all flawed ³⁰. Fifteen registered interventional trials in all, most dental, none with a phase II/III result in a systemic disease ⁸. Registered trials in malaria: none ^8,95. One in diabetes, of 80 people ⁹⁸. One in COVID-19, with eleven events ⁹⁷. The preclinical pile, by contrast, grows yearly, and in 2025 and 2026 alone produced nimbolide against mouse atherosclerosis, mouse septic kidney injury, mouse osteoporosis and mouse acetaminophen liver injury, each with a named molecular target and none with a dose a person has taken ^{78,79,80,177,178}.

Commercial interests run through the field without always being declared. The louse shampoo trials list an author from the manufacturer ¹⁰⁹; the six-month toothpaste trial compared two products of the same company ^127,128; the COVID-19 pilot tested a 'proprietary, patent-pending' extract ⁹⁷; the 2021 toxicity review has an author from the World Neem Organization ⁶⁶; the Praneem trials carry the developer's name on the author list ¹⁰⁵; the 1985 pesticide was built on a government process and sold by a sequence of companies ending in the one that lost the European patent ^24,25. None of this makes a result wrong. It is the reason to read the independent trials, which are the ones in which neem came out no better than chlorhexidine held over time ⁵⁶, no better than a standard pessary and worse tolerated ¹⁰⁶, no better than water ¹²⁶, and behind nanosilver and hypochlorite ¹³⁰.

What changed, and what is still unknown

What changed over the 58 years since 1968 is that a bitter seed became two different things. One is a molecule, C35H44O16, isolated, drawn, made in the laboratory after 22 years, and understood well enough to know that it starves and un-molts an insect through hormones a mammal does not have ^12,14,15,21. Even the molecule had its reversal: the structure published in 1975 did not survive, and the one accepted today took until the mid-1980s ^11,32. The other thing the seed became is an oil that was believed to be a tonic for infants and is a mitochondrial poison of the aspirin-metabolite class, first recognized in 1981 and never since contradicted ^22,71. The belief that reversed is the belief in the oil's safety; the belief that has not moved, because nobody has tested it properly, is the belief in the tree's medicinal power.

The known unknowns are specific. The molecular target of azadirachtin in the insect is still called 'putative' by the people who found it and 'yet to be fully understood' by the people who reviewed it ^15,65. Only the first steps of azadirachtin's biosynthesis are known, worked out in 2019, extended in 2023 to two complete routes to simpler limonoids, and traced in a third Meliaceae tree in 2026 ^4,57,58; the later steps to azadirachtin itself are not, which is why the pesticide is still extracted from trees and why a research purchase of it in 2010 held half what the label said ⁶¹. There is no human pharmacokinetic study of any neem limonoid ⁸⁰. The dose of oil that poisons a child has never been measured against body weight, because the reports record 'a few drops' and '25 to 60 ml' rather than milligrams per kilogram ¹⁰. Whether the oil's poison is its free fatty acids, which the contraceptive work isolated as the active fraction of the seed, has been proposed but not shown ^72,88. Whether neem itself prevents caries or tooth loss, as opposed to lowering a plaque score at three weeks, has not been asked: the one trial that counted cavities used a toothpaste of neem, clove and fluoride together ^30,135. Whether neem does anything in a person with malaria has not been asked either, in any registered trial ^8,95.

The absences on this page are its own findings, not claims from any source: no PubMed-indexed report of a poisoned child since 2014, although adult reports have continued and the latest is from 2025 ³⁷; no randomized malaria trial, no contraceptive efficacy trial in fertile women, no human trial of nimbolide, gedunin or nimbidin, and no trial of a neem product in any parasite of the human gut. The animal and laboratory work that fills the gap is real, and some of it, the kissing-bug rescue, the Hsp90 screen, the mouse transmission block, is elegant ^16,61,75. But a listener who asks whether to give neem for worms, for malaria, for diabetes or for a baby's skin is asking a question that 58 years of research has declined to answer, except in one respect: do not put the oil in a child's mouth.

Where it connects

Sources

178 sources, numbered as they are cited. Every one was checked against PubMed or its publisher before it was cited here; the note under each says what it shows and what it does not.

1. 1
   Alzohairy MA. Therapeutics Role of Azadirachta indica (Neem) and Their Active Constituents in Diseases Prevention and Treatment.Evid Based Complement Alternat Med · 2016 · 2016:7382506doi:10.1155/2016/7382506 · PMID 27034694

   Open-access review read in full. Gives the botany (a tree 20-23 m tall, compound leaves of 5-15 leaflets, green drupes), the use in Ayurveda, Unani and homeopathy, the constituent list, and a short section on clinical studies including a 50-patient double-blind psoriasis trial with coal tar that this page could not locate in PubMed. Its concluding line, that neem 'modulates the numerous biological processes without any adverse effect', is contradicted by the poisoning literature it itself cites.

2. 2
   Krishnan NM, Jain P, Gupta S, Hariharan AK, Panda B. An Improved Genome Assembly of Azadirachta indica A. Juss.G3 (Bethesda) · 2016 · 6(7):1835-40doi:10.1534/g3.116.030056 · PMID 27172223

   Version 2.0 of the genome built with added long reads: 3-fold higher N50, 13.4-fold less misassembly, better mapping of terpenoid-pathway genes. Assembly statistics only.

3. 3
   Krishnan NM, Pattnaik S, Jain P, et al. A draft of the genome and four transcriptomes of a medicinal and pesticidal angiosperm Azadirachta indica.BMC Genomics · 2012 · 13:464doi:10.1186/1471-2164-13-464 · PMID 22958331

   First genome of the tree and the first of any Meliaceae: rich in adenine and thymine, few repeats, about 20,000 genes, grouped by phylogeny with Citrus sinensis of the Rutaceae. A sequencing paper; it does not identify the azadirachtin biosynthetic enzymes, only candidate terpenoid genes.

4. 4
   Hodgson H, De La Pena R, Stephenson MJ, Thimmappa R, Vincent JL, Sattely ES, Osbourn A. Identification of key enzymes responsible for protolimonoid biosynthesis in plants: Opening the door to azadirachtin production.Proc Natl Acad Sci U S A · 2019 · 116(34):17096-17104doi:10.1073/pnas.1906083116 · PMID 31371503

   Open-access paper read in full. An oxidosqualene cyclase from neem (AiOSC1) and homologs from chinaberry and sweet orange make the 30-carbon triterpene tirucalla-7,24-dien-3beta-ol, and two chinaberry cytochrome P450s (MaCYP71CD2 and MaCYP71BQ5), with their citrus homologs, carry out three oxidations that close a hemiacetal ring to give the protolimonoid melianol. It states that the prototypical limonoid scaffold has 26 carbons, formed from a 30-carbon triterpene by loss of four carbons with furan-ring formation, and that the 2007 synthesis of azadirachtin took 71 steps at 0.00015% yield, so the pesticide is still obtained by extracting seed. Enzymology in yeast and tobacco; the later steps to azadirachtin remain unknown.

5. 5
   Kumar VS, Navaratnam V. Neem (Azadirachta indica): prehistory to contemporary medicinal uses to humankind.Asian Pac J Trop Biomed · 2013 · 3(7):505-14doi:10.1016/S2221-1691(13)60105-7 · PMID 23835719

   A historical review tracing neem through palm-leaf manuscripts and Ayurvedic texts to the present; indexed under smallpox, skin diseases and diabetes among its traditional uses. Narrative, not evidential; the abstract's claim that 80% of the developing world relies on traditional medicine is a WHO estimate it repeats.

6. 6
   Subapriya R, Nagini S. Medicinal properties of neem leaves: a review.Curr Med Chem Anticancer Agents · 2005 · 5(2):149-56doi:10.2174/1568011053174828 · PMID 15777222

   Review stating that neem is used in Ayurveda, Unani and homeopathy, that more than 140 compounds have been isolated from the tree, and listing claimed leaf activities. It lists claims; it does not grade them.

7. 7
   Rupani R, Chavez A. Medicinal plants with traditional use: Ethnobotany in the Indian subcontinent.Clin Dermatol · 2018 · 36(3):306-309doi:10.1016/j.clindermatol.2018.03.005 · PMID 29908572

   Dermatology review placing neem among the plants of Ayurveda, Unani, Siddha and the other traditional systems of the subcontinent. Context only.

8. 8
   Sharma A, Raut SS, Shukla A, Parul N, Singh A, Mishra A. Ethnopharmacological landscape of Azadirachta indica (Neem): Phytochemical diversity bridging traditional heritage and modern medicine.J Ethnopharmacol · 2026 · 371:122088doi:10.1016/j.jep.2026.122088 · PMID 42320775

   2026 review of literature to November 2025: over 140 compounds, preclinical activity against cancer cell lines and in streptozotocin-diabetic rodents, and the key sentence for this page: 'only fifteen interventional trials are registered, most for oro-dental conditions, with no published phase II/III results for systemic diseases', with toxicology confirming 'pediatric encephalopathy from neem oil ingestion and embryotoxicity'. Narrative review.

9. 9
   Tan QG, Luo XD. Meliaceous limonoids: chemistry and biological activities.Chem Rev · 2011 · 111(11):7437-522doi:10.1021/cr9004023 · PMID 21894902

   The standard chemical review of the limonoid family to which azadirachtin, nimbin, nimbidin, gedunin, salannin and nimbolide belong. No abstract on PubMed; cited for the classification.

10. 10
    Mishra A, Dave N. Neem oil poisoning: Case report of an adult with toxic encephalopathy.Indian J Crit Care Med · 2013 · 17(5):321-2doi:10.4103/0972-5229.120330 · PMID 24339648

    Open-access case read in full: a 73-year-old man with type 2 diabetes and psoriasis who drank 20 ml of neem oil by accident, vomited, became drowsy, convulsed in the emergency department, was intubated, had a metabolic acidosis (pH 7.34, bicarbonate 16) with normal lactate, liver and kidney tests, and recovered in four days. The paper gives the formula of azadirachtin as C35H44O16, the azadirachtin content of neem oil as 300 to over 2,500 parts per million, and summarizes earlier series: 22 infants (Lai), 12 children with 10 deaths (Sundaravalli), 13 children (Sinniah). One case; its mechanism paragraph is a hypothesis.

11. 11
    Morgan ED. Azadirachtin, a scientific gold mine.Bioorg Med Chem · 2009 · 17(12):4096-105doi:10.1016/j.bmc.2008.11.081 · PMID 19112026

    By the co-discoverer. States that the structure required 18 years to solve and the synthesis another 22, that the molecule is a feeding deterrent for many insects and a growth disruptant for most, that its mode of action, structure-activity relationships and biosynthesis still need research, that it has very low toxicity for vertebrates, and that 40 years after discovery it is still not licensed in many countries. A personal review.

12. 12
    Veitch GE, Boyer A, Ley SV. The azadirachtin story.Angew Chem Int Ed Engl · 2008 · 47(49):9402-29doi:10.1002/anie.200802675 · PMID 19031481

    The Cambridge group's own account: azadirachtin isolated from the neem tree in 1968, over 1000 publications since, and a survey of the worldwide synthesis efforts. The structural history is in the full paper; the abstract gives the 1968 date and the publication count.

13. 13
    Veitch GE, Beckmann E, Burke BJ, Boyer A, Maslen SL, Ley SV. Synthesis of azadirachtin: a long but successful journey.Angew Chem Int Ed Engl · 2007 · 46(40):7629-32doi:10.1002/anie.200703027 · PMID 17665403

    The first of the two 2007 communications reporting the first total synthesis. No abstract on PubMed; title and bibliographic data only.

14. 14
    Ley SV, Abad-Somovilla A, Anderson JC, et al. The synthesis of azadirachtin: a potent insect antifeedant.Chemistry · 2008 · 14(34):10683-704doi:10.1002/chem.200801103 · PMID 18821532

    The full paper, with 46 authors spanning two decades of work: a convergent route joining decalin and propargylic mesylate fragments, a Claisen rearrangement to build the central C8-C14 bond, and a radical cyclization for the [3.2.1] bicycle. Chemistry only; no biology.

15. 15
    Lin M, Yang S, Huang J, Zhou L. Insecticidal Triterpenes in Meliaceae: Plant Species, Molecules and Activities: Part I (Aglaia-Chukrasia).Int J Mol Sci · 2021 · 22(24):13262doi:10.3390/ijms222413262 · PMID 34948062

    Open-access review read in full. Covers 102 insecticidal triterpenes; for azadirachtin it states activity against more than 400 insect species, that the primary antifeedant effect is mediated by gustatory chemosensilla, that azadirachtin is an antagonist of 20-hydroxyecdysone and juvenile hormone acting through inhibition of prothoracicotropic hormone release and of ecdysone 20-monooxygenase, that it damages the prothoracic gland, corpus allatum and corpus cardiacum, binds heat-shock protein 60, and that 'the exact molecular mechanism of action of azadirachtin has yet to be fully understood'. Review of insect work.

16. 16
    Garcia ES, Feder D, Gomes JE, de Azambuja P. Effects of precocene and azadirachtin in Rhodnius prolixus: some data on development and reproduction.Mem Inst Oswaldo Cruz · 1987 · 82 Suppl 3:67-73doi:10.1590/s0074-02761987000700014 · PMID 3334403

    In the kissing bug, azadirachtin inhibited molting and reproduction; the molting block was fully reversed by giving ecdysone; ecdysteroid titers fell; in vitro data suggested direct action on the prothoracic glands and ovaries. Insect work.

17. 17
    Smith SL, Mitchell MJ. Effects of azadirachtin on insect cytochrome P-450 dependent ecdysone 20-monooxygenase activity.Biochem Biophys Res Commun · 1988 · 154(2):559-63doi:10.1016/0006-291x(88)90176-3 · PMID 2900005

    Azadirachtin inhibited ecdysone 20-monooxygenase in homogenates of Drosophila, Aedes and Manduca, with 50% inhibition at 1 x 10^-4 M (Drosophila) to 4 x 10^-4 M (Manduca midgut). In vitro, at concentrations far above those active on whole insects.

18. 18
    European Food Safety Authority; Arena M, Auteri D, Barmaz S, et al. Peer review of the pesticide risk assessment of the active substance azadirachtin (Margosa extract).EFSA J · 2018 · 16(9):e05234doi:10.2903/j.efsa.2018.5234 · PMID 32626038

    Open-access regulatory peer review read in full. Azadirachtin was approved in the European Union on 1 June 2011, with a specific provision that only insecticide uses be authorized; this review covers an application by the manufacturer to extend the approval to use as an acaricide on greenhouse ornamentals. The reference values are expressed as whole extract, not as azadirachtin A: acceptable daily intake and acceptable operator exposure level 0.1 mg/kg body weight per day, from a 90-day rat study with a no-observed-adverse-effect level of 32 mg/kg per day and the liver as target organ, divided by a safety factor of 300; acute reference dose 0.75 mg/kg from a rat developmental study with a maternal no-effect level of 225 mg/kg per day. The three commercial extracts were clastogenic in cultured human lymphocytes but not in animals, and fertility and reproductive performance were unimpaired in a valid multigeneration rat study. Data gaps remain for long-term and carcinogenicity studies, a rabbit developmental study, bioavailability, and the fate of the other limonoids and the non-limonoid components of the extract.

19. 19
    Raizada RB, Srivastava MK, Kaushal RA, Singh RP. Azadirachtin, a neem biopesticide: subchronic toxicity assessment in rats.Food Chem Toxicol · 2001 · 39(5):477-83doi:10.1016/s0278-6915(00)00153-8 · PMID 11313114

    Technical azadirachtin (12%) given by mouth to rats at 500, 1,000 and 1,500 mg/kg/day for 90 days produced no sign of toxicity, death, organ-weight change, pathology or blood change; the authors propose 1,500 mg/kg as the base for a no-observed-effect level. Note the test material was 12% azadirachtin, so the pure-molecule dose was about an eighth of the stated figure. Rats.

20. 20
    Boeke SJ, Boersma MG, Alink GM, et al. Safety evaluation of neem (Azadirachta indica) derived pesticides.J Ethnopharmacol · 2004 · 94(1):25-41doi:10.1016/j.jep.2004.05.011 · PMID 15261960

    Review of oral toxicity data in humans and animals. Non-aqueous extracts were the most toxic (estimated safe dose 0.002 to 12.5 microg/kg/day); seed oil and aqueous extracts less so (0.26-0.3 mg/kg/day and 2 microl/kg/day); pure azadirachtin least (15 mg/kg/day). Reversible effects on reproduction in both sexes were the most important sub-acute and chronic effects. The two non-aqueous figures are two separate estimates, not a range. A risk-assessment review from an entomology laboratory.

21. 21
    Scott RH, O'Brien K, Roberts L, Mordue W, Mordue Luntz J. Extracellular and intracellular actions of azadirachtin on the electrophysiological properties of cultured rat DRG neurones.Comp Biochem Physiol C Pharmacol Toxicol Endocrinol · 1999 · 123(1):85-93doi:10.1016/s0742-8413(99)00014-6 · PMID 10390060

    Rat sensory neurons were 'relatively insensitive': insect preparations respond at 1-100 nM, the rat cells needed 10-100 microM, at which potassium currents were reversibly altered. Cultured rat cells; the clearest head-to-head on why the molecule spares mammals.

22. 22
    Sinniah D, Baskaran G. Margosa oil poisoning as a cause of Reye's syndrome.Lancet · 1981 · 1(8218):487-9doi:10.1016/s0140-6736(81)91861-4 · PMID 6110100

    The founding report: thirteen infants in Malaysia developed vomiting, drowsiness, metabolic acidosis, neutrophil leukocytosis and encephalopathy within hours of being given margosa oil; liver biopsy in one and necropsy of experimentally poisoned mice showed fatty infiltration of liver and proximal kidney tubules, cerebral edema and mitochondrial damage. A case series with an animal check; no dose data in the abstract, and the features are listed for the group of thirteen rather than stated for each infant.

23. 23
    Sundaravalli N, Raju BB, Krishnamoorthy KA. Neem oil poisoning.Indian J Pediatr · 1982 · 49(398):357-9doi:10.1007/BF02834422 · PMID 7141505

    The Indian series, indexed under infant, preschool child, liver and seizures; no abstract on PubMed. Mishra and Dave summarize it as 12 children given single doses of 25-60 ml, 10 of whom died with encephalopathy and metabolic acidosis.

24. 24
    National Research Council. Neem: A Tree for Solving Global Problems. Appendix A.National Academies Press · Washington, DC · 1992

    Primary document, not indexed in PubMed. Reached through a web-search summary; the page itself could not be opened from this session, so only the statements the summary reproduced are used: that the Environmental Protection Agency approved a commercial neem-based insecticide, Margosan-O, an ethanolic seed-extract concentrate standardized to 3,000 ppm azadirachtin (plus or minus 10%) and made by a USDA process, for certain nonfood uses in 1985, the registrant being Vikwood, Ltd. of Sheboygan, Wisconsin. EPA label records surfaced by the same search show the product line passing to W. R. Grace (a 1992 label) and Thermo Trilogy (a 1999 label). Every statement on this page that rests on this source alone is attributed in the text to the 1992 National Research Council report.

25. 25
    European Patent Office, Technical Board of Appeal. Decision T 0416/01 of 8 March 2005 on European patent 0436257.EPO · 2005 · case T 0416/01

    Primary document, not indexed in PubMed. Reached through a web-search summary; the decision PDF could not be opened from this session. The decision itself was not opened from this session, and its date and outcome are corroborated by the contemporaneous reports in Nature and Nature Biotechnology [176][67]. The summary reports: patent publication number 0436257, proprietor Thermo Trilogy Corporation, the main request found to lack an inventive step because it was obvious to try such formulations, the auxiliary request rejected for added matter, and the 2000 revocation left standing. Press reports summarized alongside it give the first revocation as 11 May 2000 and describe the original grantees as the United States Department of Agriculture and W. R. Grace; the board's own text names only Thermo Trilogy.

26. 26
    Sheridan C. EPO neem patent revocation revives biopiracy debate.Nat Biotechnol · 2005 · 23(5):511-2doi:10.1038/nbt0505-511 · PMID 15877054

    News item, no abstract on PubMed; the title records that the European Patent Office's revocation of the neem patent was upheld in 2005 and reopened the biopiracy argument.

27. 27
    Hamilton C. Biodiversity, biopiracy and benefits: what allegations of biopiracy tell us about intellectual property.Dev World Bioeth · 2006 · 6(3):158-73doi:10.1111/j.1471-8847.2006.00168.x · PMID 17038007

    Uses the neem tree case as the worked example of a biopiracy allegation, arguing that activists deliberately simplified the science and patent law to challenge the intellectual-property system itself. A bioethics essay.

28. 28
    Hellerer U, Jarayaman KS. Greens persuade Europe to revoke patent on neem tree.Nature · 2000 · 405(6784):266-7doi:10.1038/35012778 · PMID 10830934

    News item, no abstract on PubMed, indexed under antifungal agents, patents, India and Europe; cited for the fact and the date of the European Patent Office's revocation of the neem patent in May 2000.

29. 29
    Dickson D, Jayaraman KS. Aid groups back challenge to neem patents.Nature · 1995 · 377(6545):95doi:10.1038/377095a0 · PMID 7675092

    News item, no abstract on PubMed, indexed under patents, insecticides, plant oils, India, Europe and the United States; cited for the date at which aid organizations backed the organized challenge to the neem patents.

30. 30
    Dhingra K, Vandana KL. Effectiveness of Azadirachta indica (neem) mouthrinse in plaque and gingivitis control: a systematic review.Int J Dent Hyg · 2016 · 15(1):4-15doi:10.1111/idh.12191 · PMID 26876277

    Systematic review to February 2015: of 206 articles, three randomized trials qualified; too heterogeneous for meta-analysis; each reported neem as effective as chlorhexidine as an adjunct to brushing, but 'the quality of reporting and evidence along with methods of studies was generally flawed with unclear risk of bias'. The honest summary of the field.

31. 31
    Meeran M, Murali A, Balakrishnan R, Narasimhan D. 'Herbal remedy is natural and safe'--truth or myth?J Assoc Physicians India · 2013 · 61(11):848-50PMID 24974507

    An adult with neem oil toxicity: severe metabolic acidosis, seizures, kidney failure and encephalopathy after a small dose, treated with fluids. One case; the title is the authors' own verdict.

32. 32
    Zanno P, Miura I, Nakanishi K, Elder DL. Letter: Structure of the insect phagorepellent azadirachtin. Application of PRFT/CWD carbon-13 nuclear magnetic resonance.J Am Chem Soc · 1975 · 97(7):1975-7doi:10.1021/ja00840a073 · PMID 1133406

    The first published structure of azadirachtin, derived from carbon-13 nuclear magnetic resonance. No abstract in PubMed, so it is cited only for its title, date and method. Morgan's statement that the structure 'required 18 years to solve' [7] places the accepted structure in the mid-1980s, which is why this page treats the 1975 proposal as one that did not stand. The 1985-1987 corrections by Kraus and by Broughton and Ley are not indexed in PubMed and were not opened from this session, so they are not cited.

33. 33
    Lai SM, Lim KW, Cheng HK. Margosa oil poisoning as a cause of toxic encephalopathy.Singapore Med J · 1990 · 31(5):463-5PMID 2259944

    Two Singapore infants; the abstract describes the usual picture (vomiting, drowsiness, rapid breathing, recurrent generalized seizures, leukocytosis, metabolic acidosis), notes the oil is 'often administered orally to neonates and infants regularly in small amounts', and that prognosis is usually good but deaths and neurological deficits occur. Case reports; the abstract describes leukocytosis and metabolic acidosis as 'significant laboratory findings' rather than as universal ones.

34. 34
    Senanayake MP, Rupasinghe S, Dissanayake PV. Margosa (Kohomba) oil induced toxic encephalopathy following home remedy for intestinal worms.Ceylon Med J · 2009 · 54(4):140doi:10.4038/cmj.v54i4.1459 · PMID 20052860

    Letter, no abstract: a Sri Lankan infant given the oil as a worm remedy developed toxic encephalopathy. Included because the indication was intestinal worms.

35. 35
    Kumar S, Kumar N. Neem oil poisoning as a cause of toxic encephalopathy in an infant.Indian J Pediatr · 2014 · 81(9):955doi:10.1007/s12098-013-1327-x · PMID 24522945

    Letter, no abstract: an infant with toxic encephalopathy after neem oil, 2014. The most recent PubMed-indexed pediatric case this page found.

36. 36
    Sinniah D, Baskaran G, Looi LM, Leong KL. Reye-like syndrome due to margosa oil poisoning: report of a case with postmortem findings.Am J Gastroenterol · 1982 · 77(3):158-61PMID 7081175

    A fatal newborn case with autopsy of liver and kidney. No abstract on PubMed; indexed under Reye syndrome and aflatoxin poisoning, the latter being the differential the authors had to exclude.

37. 37
    Alla KR, Pulluri R. Hypernatremia Caused by Saltwater Emesis in a Case of Neem Powder Poisoning.Cureus · 2025 · 17(8):e89685doi:10.7759/cureus.89685 · PMID 40932984

    A young woman with suspected neem powder self-poisoning was made to vomit with salt water at home and reached a serum sodium of 187 mmol/L; she recovered completely with treatment. One case, and the injury described came from the home remedy rather than from the plant; the authors call for the practice to be abandoned.

38. 38
    Gathirwa JW, Rukunga GM, Mwitari PG, et al. Traditional herbal antimalarial therapy in Kilifi district, Kenya.J Ethnopharmacol · 2011 · 134(2):434-42doi:10.1016/j.jep.2010.12.043 · PMID 21211554

    Methanolic neem extract was among the most active of 13 Kenyan antimalarial plants in vitro (IC50 under 10 microg/ml against both chloroquine-sensitive and resistant clones) and gave 89% parasite suppression in mice. In vitro and mouse.

39. 39
    MacKinnon S, Durst T, Arnason JT, et al. Antimalarial activity of tropical Meliaceae extracts and gedunin derivatives.J Nat Prod · 1997 · 60(4):336-41doi:10.1021/np9605394 · PMID 9134742

    Of 22 Meliaceae extracts, twelve were active against Plasmodium falciparum in vitro, including neem leaves, whose active limonoid was gedunin; only gedunin beat chloroquine against the chloroquine-resistant W2 clone. In vitro. Note the clone labels: the abstract calls W2 chloroquine-sensitive and D6 chloroquine-resistant, the reverse of the standard designations of those two laboratory clones, and this page reports the comparison as the paper states it.

40. 40
    Udeinya IJ, Brown N, Shu EN, Udeinya FI, Quakeyie I. Fractions of an antimalarial neem-leaf extract have activities superior to chloroquine, and are gametocytocidal.Ann Trop Med Parasitol · 2006 · 100(1):17-22doi:10.1179/136485906X78508 · PMID 16417709

    Two leaf fractions inhibited asexual P. falciparum at nanogram concentrations and lysed 50% of gametocytes at 10^-3 microg/ml, with the authors' own condition: 'if they are found safe and effective in vivo'. In vitro; no in vivo follow-up located.

41. 41
    Ohouko OFH, Lègba BB, Agbodjento E, et al. Diabetes mellitus in ECOWAS: an overview of the safety and efficacy of medicinal plants used in traditional medicine practices.Trop Med Health · 2026 · 54(1)doi:10.1186/s41182-026-00936-4 · PMID 41872917

    2026 review: neem is among the twelve most-cited antidiabetic plants in West Africa, but 'few of these plants have undergone clinical trials and in-depth toxicological evaluations'. Review.

42. 42
    Gitonga D, Hajihassani A. A novel nematicide application method based on nematode life cycle for managing root-knot nematode, Meloidogyne incognita.Pest Manag Sci · 2025 · 82(2):2081-2089doi:10.1002/ps.70350 · PMID 41201020

    In greenhouse tomato, azadirachtin and thyme oil reduced root galling and reproduction of the root-knot nematode; in field cucumber, azadirachtin was the one calendar-applied product that reduced nematode density. Plant-parasitic nematodes in soil, not animal parasites.

43. 43
    Sánchez-Gilo A, Nuño González A, Gutiérrez Pascual M, Vicente Martín FJ. Airborne Allergic Contact Dermatitis Caused by Neem Oil.Actas Dermosifiliogr (Engl Ed) · 2018 · 109(5):449-450doi:10.1016/j.ad.2017.05.020 · PMID 29074140

    Letter: a middle-aged man with airborne allergic contact dermatitis from neem oil. One case.

44. 44
    Bernaola M, Valls A, de Frutos C, Garcia-Abujeta JL. Occupational allergic contact dermatitis to neem oil used in natural cosmetic.Contact Dermatitis · 2020 · 82(6):389-390doi:10.1111/cod.13490 · PMID 32043597

    Case report of occupational facial and forearm dermatitis in a woman making natural cosmetics with neem oil, confirmed by patch test. One case.

45. 45
    Samaran Q, Dereure O, Raison-Peyron N. Allergic contact dermatitis to neem oil used to treat a flare of atopic dermatitis.Contact Dermatitis · 2023 · 89(5):385-387doi:10.1111/cod.14394 · PMID 37557893

    Case report: neem oil applied to eczema caused allergic contact dermatitis. One case, 2023.

46. 46
    El Kaderi Y, Kerre S, Herman A. Allergy to Neem Oil: An Emerging Cause of Contact Dermatitis.Contact Dermatitis · 2025 · 93(4):339-340doi:10.1111/cod.14832 · PMID 40571349

    Case report, 2025, from Belgium, calling neem oil in cosmetics 'an emerging cause' of contact allergy. One case.

47. 47
    Lauriola MM, Corazza M. Allergic contact dermatitis caused by argan oil, neem oil, and Mimosa tenuiflora.Contact Dermatitis · 2016 · 75(6):388-390doi:10.1111/cod.12665 · PMID 27870098

    Case report of a psoriasis patient sensitized to three herbal oils including neem. One case; the indication was psoriasis.

48. 48
    de Groot A, Jagtman BA, Woutersen M. Contact Allergy to Neem Oil.Dermatitis · 2017 · 28(6):360-362doi:10.1097/DER.0000000000000309 · PMID 29059091

    One case of allergic contact dermatitis from neem oil, with the authors' note that contact allergy to the oil had been described in only three patients before and that the allergen is unknown.

49. 49
    Romita P, Calogiuri G, Bellino M, De Prezzo S, Ambrogio F, Foti C. Allergic contact dermatitis caused by neem oil: An underrated allergen?Contact Dermatitis · 2019 · 81(2):133-134doi:10.1111/cod.13256 · PMID 30834541

    An Italian case confirmed by patch testing, in a patient also reacting to argan and olive oils. No abstract in PubMed; cited for the case, its date and its patch-test confirmation.

50. 50
    Tamagawa-Mineoka R, Masuda K, Katoh N. Allergic contact dermatitis due to neem oil: A case report and mini-review.J Dermatol · 2020 · 47(2):e48-e49doi:10.1111/1346-8138.15182 · PMID 31823408

    A Japanese case in a woman who applied neem oil to the face for acne, confirmed by patch testing. No abstract in PubMed; indexed under allergic contact dermatitis, patch tests and Ayurvedic medicine.

51. 51
    Ambooken B, Abdulsalam S, Asokan N, Abraham A. Allergic contact stomatitis caused by neem leaves (Azadirachta indica).Contact Dermatitis · 2017 · 76(6):376-378doi:10.1111/cod.12732 · PMID 28497627

    An allergic reaction of the mouth lining caused by chewing neem leaves. One case; no abstract in PubMed.

52. 52
    Dhongade RK, Kavade SG, Damle RS. Neem oil poisoning.Indian Pediatr · 2008 · 45(1):56-7PMID 18250509

    A previously normal 5-year-old with refractory seizures and metabolic acidosis after neem oil, indexed under cardiac arrest and status epilepticus, left with hearing and visual disturbance and ataxia. One case; the clearest report of permanent damage.

53. 53
    Iyyadurai R, Surekha V, Sathyendra S, Paul Wilson B, Gopinath KG. Azadirachtin poisoning: a case report.Clin Toxicol (Phila) · 2010 · 48(8):857-8doi:10.3109/15563650.2010.518148 · PMID 20932102

    A 35-year-old woman who drank a neem-kernel pesticide in self-harm developed neurotoxicity needing ventilation and survived without lasting effects; the authors note that adult overdose of these 'natural and safe' pesticides had not been described. One case; the product was a formulation, so the dose of azadirachtin itself is unknown.

54. 54
    Bhaskar MV, Pramod SJ, Jeevika MU, Chandan PK, Shetteppa G. MR imaging findings of neem oil poisoning.AJNR Am J Neuroradiol · 2010 · 31(7):E60-1doi:10.3174/ajnr.A2146 · PMID 20448012

    Letter, no abstract, indexed under adult female, attempted suicide and brain disease; Mishra and Dave summarize it as a 35-year-old woman with bilateral visual loss and symmetrical signal change in both putamina extending into the internal capsule, who recovered completely with normal laboratory findings. One case; neither the letter nor its summary reports follow-up imaging.

55. 55
    Suresha AR, Rajesh P, Anil Raj KS, Torgal R. A rare case of toxic optic neuropathy secondary to consumption of neem oil.Indian J Ophthalmol · 2014 · 62(3):337-9doi:10.4103/0301-4738.121129 · PMID 24722271

    A 35-year-old woman presented with two days of bilateral visual loss five days after drinking about 150 ml of neem oil: no perception of light in either eye, dilated sluggish pupils, hyperemic swollen discs, and altered signal in both putaminal regions on MRI read as cytotoxic edema from tissue hypoxia; vision reached 20/200 in both eyes at two months. One case, and possibly the same patient as the 2010 imaging letter [59].

56. 56
    Sharma R, Hebbal M, Ankola AV, Murugaboopathy V, Shetty SJ. Effect of two herbal mouthwashes on gingival health of school children.J Tradit Complement Med · 2014 · 4(4):272-8doi:10.4103/2225-4110.131373 · PMID 25379471

    105 children aged 12-15 randomized to neem, mango or 0.2% chlorhexidine mouthwash for 21 days and followed three months: all three reduced plaque and gingivitis; chlorhexidine alone kept the plaque reduction at two months and had 'superior antigingivitis properties'. The largest neem mouthwash trial found; chlorhexidine won on durability.

57. 57
    De La Pena R, Hodgson H, Liu JC, et al. Complex scaffold remodeling in plant triterpene biosynthesis.Science · 2023 · 379(6630):361-368doi:10.1126/science.adf1017 · PMID 36701471

    Open-access paper read in full. Twenty-two enzymes from chinaberry and citrus, including two sterol isomerases given a new function, catalyze twelve reactions that complete the biosynthesis of the furan-bearing limonoids azadirone and kihadalactone A. It notes that azadirachtin is active against more than 600 insect species and is available commercially only as a component of neem oil, and that the ring-C cleavage steps leading to azadirachtin itself are still to be found. Enzymology in tobacco.

58. 58
    Liu B, Feng L, Zhang K, et al. Adaptive evolution of oxidosqualene cyclases has driven diversification of triterpene scaffolds for limonoid biosynthesis in Melia toosendan.Plant Commun · 2026 · 101965doi:10.1016/j.xplc.2026.101965 · PMID 42289993

    2026: a chromosome-level genome of Melia toosendan with 14 predicted oxidosqualene cyclases, 11 of which made distinct triterpene skeletons when expressed in tobacco, and four amino-acid substitutions that switch the product of the enzyme making the canonical limonoid precursor. Plant molecular biology; it concerns a relative of neem, not neem itself.

59. 59
    R TK, Bhat MDA, Zaman R, Najar FA. Efficacy of herbal anti-microbial soap in Tinea corporis: A randomized controlled study.J Ethnopharmacol · 2022 · 287:114934doi:10.1016/j.jep.2021.114934 · PMID 34968662

    Unani institute trial: a soap of turmeric, neem and Cassia tora against placebo soap in 30 patients with ringworm for four weeks; KOH mount turned negative in 80% versus 20%. Single-blind, 15 per arm, three plants.

60. 60
    Kaur G, Sarwar Alam M, Athar M. Nimbidin suppresses functions of macrophages and neutrophils: relevance to its antiinflammatory mechanisms.Phytother Res · 2004 · 18(5):419-24doi:10.1002/ptr.1474 · PMID 15174005

    Nimbidin, described as a mixture of tetranortriterpenes and the major active principle of the seed oil, at 5-25 mg/kg by mouth for 3 days in rats inhibited macrophage migration, phagocytosis and respiratory burst, and in vitro cut nitric oxide and prostaglandin E2 output and neutrophil degranulation. Rats and rat cells.

61. 61
    Lucantoni L, Yerbanga RS, Lupidi G, Pasqualini L, Esposito F, Habluetzel A. Transmission blocking activity of a standardized neem (Azadirachta indica) seed extract on the rodent malaria parasite Plasmodium berghei in its vector Anopheles stephensi.Malar J · 2010 · 9:66doi:10.1186/1475-2875-9-66 · PMID 20196858

    Open-access paper read in full. NeemAzal (34% azadirachtin A, 56% total limonoids) injected into gametocytemic mice at 50 mg/kg azadirachtin one hour before mosquito feeding blocked oocyst formation in all 138 mosquitoes and none of six challenged mice became infected; 25 and 13.2 mg/kg did nothing, and pure azadirachtin at a nominal 50 mg/kg did nothing, which the authors trace to the commercial azadirachtin containing half the declared amount. Mouse model; the discussion notes limonoids' short half-life and poor absorption.

62. 62
    Tufail T, Bader Ul Ain H, Ijaz A, et al. Neem (Azadirachta indica): A Miracle Herb; Panacea for All Ailments.Food Sci Nutr · 2025 · 13(9):e70820doi:10.1002/fsn3.70820 · PMID 40901661

    A 2025 review whose title says 'miracle herb' and 'panacea for all ailments' and whose abstract calls the tree 'a natural remedy for numerous illnesses'. Cited as a specimen of the tone that the evidence on this page does not support; it does list the constituents and the Ayurvedic, Unani and homeopathic uses.

63. 63
    Mitchell MJ, Smith SL, Johnson S, Morgan ED. Effects of the neem tree compounds azadirachtin, salannin, nimbin, and 6-desacetylnimbin on ecdysone 20-monooxygenase activity.Arch Insect Biochem Physiol · 1997 · 35(1-2):199-209doi:10.1002/(SICI)1520-6327(1997)35:1/2<199::AID-ARCH18>3.0.CO;2-6 · PMID 9131784

    All four seed limonoids inhibited the enzyme that activates ecdysone, with 50% inhibition between about 2 x 10^-5 and 1 x 10^-3 M across three insect species. In vitro.

64. 64
    Nisbet AJ, Mordue Luntz AJ, Grossman RB, Jennens L, Ley SV, Mordue W. Characterization of azadirachtin binding to Sf9 nuclei in vitro.Arch Insect Biochem Physiol · 2001 · 46(1-2):78-86doi:10.1002/arch.11 · PMID 11276064

    Tritiated dihydroazadirachtin bound specifically and essentially irreversibly to a single class of sites in insect cell nuclei, Kd 18.1 nM, Bmax 23.87 pmol/mg protein; azadirachtin itself competed with IC50 1.55 x 10^-8 M. Republication of 1997 data. Insect cells.

65. 65
    Robertson SL, Ni W, Dhadialla TS, et al. Identification of a putative azadirachtin-binding complex from Drosophila Kc167 cells.Arch Insect Biochem Physiol · 2007 · 64(4):200-8doi:10.1002/arch.20171 · PMID 17366600

    Azadirachtin halved proliferation of Drosophila cells at 1.7 x 10^-7 M over 48 h, caused nuclear damage within 6 h at 5 x 10^-7 M, and a binding complex pulled from the cells contained heat-shock protein 60. 'Putative' is the authors' word. Insect cells.

66. 66
    Braga TM, Rocha L, Chung TY, et al. Azadirachta indica A. Juss. In Vivo Toxicity-An Updated Review.Molecules · 2021 · 26(2):252doi:10.3390/molecules26020252 · PMID 33419112

    Open-access review read in full. Collects acute, sub-acute, sub-chronic and reproductive toxicity studies: oral LD50 of neem oil in mice 31.95 g/kg; a 90-day mouse study giving 177 mg/kg/day as the no-observed-adverse-effect level with kidney, liver and testis as target organs at 1,600 mg/kg; a two-generation rat study of dietary azadirachtin (5-50 mg/kg) without reproductive or teratogenic effect; male-rat sperm and testis changes with leaf powder, leaf extracts, bark extract, azadirachtin-A and nimbolide; an intramuscular LD50 for leaf and seed water extracts of 6.2 and 9.4 ml/kg; and azadirachtin chromosome damage in a fish. One author is affiliated with the World Neem Organization, a commercial conflict the paper discloses.

67. 67
    Nisbet AJ, Mordue Luntz AJ, Williams LM, et al. Autoradiographic localization of [22,23-3H2] dihydroazadirachtin binding sites in desert locust testes and effects of azadirachtin on sperm motility.Tissue Cell · 1996 · 28(6):725-9doi:10.1016/s0040-8166(96)80075-9 · PMID 18621343

    Labeled azadirachtin bound to developing locust sperm tails at 10^-8 M; motility of locust sperm fell at 10^-4 M and above, and boar sperm forward motility fell above 10^-5 M. The boar result is the one mammalian cell effect at a defined concentration on this page; in vitro.

68. 68
    Billker O, Shaw MK, Jones IW, Ley SV, Mordue AJ, Sinden RE. Azadirachtin disrupts formation of organised microtubule arrays during microgametogenesis of Plasmodium berghei.J Eukaryot Microbiol · 2002 · 49(6):489-97doi:10.1111/j.1550-7408.2002.tb00234.x · PMID 12503686

    In the rodent malaria parasite, azadirachtin left DNA replication and tubulin assembly intact but blocked the organization of mitotic spindles and axonemes, so motile male gametes were not released; it did not inhibit growth of asexual blood stages. Microscopy of cultured parasites.

69. 69
    Dallaqua B, Saito FH, Rodrigues T, et al. Azadirachta indica treatment on the congenital malformations of fetuses from rats.J Ethnopharmacol · 2013 · 150(3):1109-13doi:10.1016/j.jep.2013.10.046 · PMID 24184262

    Pregnant rats given neem seed oil (1.2 ml/day) throughout pregnancy had more malformed or variant fetuses, especially visceral; azadirachtin (1.0 mg/ml/day) did not; neither changed maternal glucose. Rats; the cleanest separation of oil from molecule on this page.

70. 70
    Koga Y, Yoshida I, Kimura A, Yoshino M, Yamashita F, Sinniah D. Inhibition of mitochondrial functions by margosa oil: possible implications in the pathogenesis of Reye's syndrome.Pediatr Res · 1987 · 22(2):184-7doi:10.1203/00006450-198708000-00017 · PMID 3658544

    Isolated rat liver mitochondria: margosa oil is an uncoupler, raising state 4 respiration, lowering the respiratory control ratio and ATP content and shifting coenzyme A esters; coenzyme Q and carnitine partly relieved the block on electron transport. In vitro.

71. 71
    Trost LC, Lemasters JJ. The mitochondrial permeability transition: a new pathophysiological mechanism for Reye's syndrome and toxic liver injury.J Pharmacol Exp Ther · 1996 · 278(3):1000-5PMID 8819478

    Neem oil, salicylate, valproate and the other Reye-associated chemicals all opened the cyclosporin-sensitive permeability pore of isolated rat liver mitochondria; cyclosporin A blocked it. In vitro; places neem oil in the same mechanistic class as aspirin's metabolite.

72. 72
    Sinniah R, Sinniah D, Chia LS, Baskaran G. Animal model of margosa oil ingestion with Reye-like syndrome. Pathogenesis of microvesicular fatty liver.J Pathol · 1989 · 159(3):255-64doi:10.1002/path.1711590313 · PMID 2593049

    Time course in mice: nuclear changes within 30 minutes, mitochondrial swelling and loss of ribosomes from 60 minutes, glycogen loss, proliferation of endoplasmic reticulum, then lipid droplets; the authors call margosa oil 'a long-chain fatty acid compound' and note deaths from hepatoencephalopathy in children in Malaysia and India. Mice.

73. 73
    Sinniah D, Schwartz PH, Mitchell RA, Arcinue EL. Investigation of an animal model of a Reye-like syndrome caused by Margosa oil.Pediatr Res · 1985 · 19(12):1346-55doi:10.1203/00006450-198512000-00029 · PMID 4080457

    Rats injected twice with margosa oil developed florid neurological signs with high ammonia and transaminases, low glucose, fatty liver with glycogen depletion and mitochondrial changes on electron microscopy, but, unlike Reye's syndrome, no fall in hepatic mitochondrial enzymes and no brain edema. Rats.

74. 74
    Pillai NR, Santhakumari G. Anti-arthritic and anti-inflammatory actions of nimbidin.Planta Med · 1981 · 43(1):59-63doi:10.1055/s-2007-971474 · PMID 7345443

    The early rat study of nimbidin in edema, granuloma and arthritis models; no abstract on PubMed. Rats.

75. 75
    Hieronymus H, Lamb J, Ross KN, et al. Gene expression signature-based chemical genomic prediction identifies a novel class of HSP90 pathway modulators.Cancer Cell · 2006 · 10(4):321-30doi:10.1016/j.ccr.2006.09.005 · PMID 17010675

    Gedunin and celastrol were found as inhibitors of androgen-receptor signaling in prostate cancer cells, and a gene-expression comparison predicted, then confirmed, that they inhibit heat-shock protein 90 and its client proteins. Cell lines; the origin of gedunin's Hsp90 story.

76. 76
    Jang SW, Liu X, Chan CB, et al. Deoxygedunin, a natural product with potent neurotrophic activity in mice.PLoS One · 2010 · 5(7):e11528doi:10.1371/journal.pone.0011528 · PMID 20644624

    Deoxygedunin, from the neem gedunin family, activated the TrkB receptor in mouse brain independently of BDNF, prevented vestibular ganglion degeneration in BDNF-deficient pups, and had antidepressant and learning effects in mice. Mice.

77. 77
    Nagini S, Palrasu M, Bishayee A. Limonoids from neem (Azadirachta indica A. Juss.) are potential anticancer drug candidates.Med Res Rev · 2024 · 44(2):457-496doi:10.1002/med.21988 · PMID 37589457

    Review of azadirachtin, epoxyazadiradione, gedunin and nimbolide in cancer cell lines and animal models: azadirachtin blocks mitotic spindle formation, gedunin inhibits Hsp90, nimbolide is 'the most potent cytotoxic neem limonoid'. All preclinical; the review reports no human trial.

78. 78
    Qin Y, Zhou Y, Liu C, Zou J, He C. Nimbolide competitively binds to the 14-3-3 epsilon with TFEB to promote lipophagy and ameliorate atherosclerosis.Phytomedicine · 2026 · 161:158774doi:10.1016/j.phymed.2026.158774 · PMID 42705181

    2026: nimbolide reduced aortic plaque in ApoE-deficient mice and was shown by thermal shift and surface plasmon resonance to bind 14-3-3 epsilon, displacing TFEB. Mice and macrophages.

79. 79
    Ba J, Lin Y, Zhang J, Wang Y, Wu B. Nimbolide Targeting SIRT1 Protects Against Acetaminophen-Induced Acute Liver Injury by Regulating Oxidative Stress and Endoplasmic Reticulum Stress.Pharmacol Res Perspect · 2025 · 13(3):e70120doi:10.1002/prp2.70120 · PMID 40375435

    2025: nimbolide 20 mg/kg intraperitoneally two hours before a toxic acetaminophen dose reduced liver injury in mice via SIRT1. Mice; pretreatment, not rescue.

80. 80
    Jogi M, Asnani H, Singh S, Kumar P. Nimbolide: A Potential Phytochemical Agent in Multimodal Pancreatic Cancer Therapies.Mini Rev Med Chem · 2025 · 25(1):27-41doi:10.2174/0113895575293138240527061556 · PMID 38874049

    2025 review stating that nimbolide 'is still in the early stages of the drug development process because no comprehensive pharmacokinetic studies or long-term toxicity studies' exist, and that preclinical toxicology must precede 'initial human clinical trials'. Review.

81. 81
    Almas K. The antimicrobial effects of extracts of Azadirachta indica (Neem) and Salvadora persica (Arak) chewing sticks.Indian J Dent Res · 1999 · 10(1):23-6PMID 10865390

    Aqueous extracts of neem and arak chewing sticks both inhibited Streptococcus mutans and Streptococcus faecalis on agar at 50% concentration; arak worked at lower concentrations against S. faecalis. Agar plates, not mouths.

82. 82
    Kankariya AR, Patel AR, Kunte SS. The effect of different concentrations of water soluble azadirachtin (neem metabolite) on Streptococcus mutans compared with chlorhexidine.J Indian Soc Pedod Prev Dent · 2016 · 34(2):105-10doi:10.4103/0970-4388.180394 · PMID 27080957

    Plaque from 30 children aged 8-12: a 40% water-soluble azadirachtin preparation inhibited S. mutans in zone and biofilm assays as well as 0.2% chlorhexidine. In vitro; note that 40% is an enormous concentration.

83. 83
    Mistry KS, Sanghvi Z, Parmar G, Shah S, Pushpalatha K. Antibacterial efficacy of Azadirachta indica, Mimusops elengi and 2% CHX on multispecies dentinal biofilm.J Conserv Dent · 2015 · 18(6):461-6doi:10.4103/0972-0707.168810 · PMID 26752840

    On a 21-day four-species biofilm grown on extracted teeth, methanolic neem extract reduced counts against saline but 2% chlorhexidine was significantly more effective than either plant extract. In vitro; a loss for neem against the standard.

84. 84
    Veitch GE, Beckmann E, Burke BJ, Boyer A, Ayats C, Ley SV. A relay route for the synthesis of azadirachtin.Angew Chem Int Ed Engl · 2007 · 46(40):7633-5doi:10.1002/anie.200703028 · PMID 17665383

    Companion communication completing the 2007 synthesis by a relay route from a natural-product-derived intermediate. No abstract on PubMed.

85. 85
    Sinniah D, Sinniah R, Baskaran G, Pathmanathan R, Yamashita F, Yoshino M. Evaluation of the possible role of glucose, carnitine, coenzyme Q10 and steroids in the treatment of Reye's syndrome using the margosa oil animal model.Acta Paediatr Jpn · 1990 · 32(4):462-8doi:10.1111/j.1442-200x.1990.tb00861.x · PMID 2288230

    In poisoned mice, glucose improved survival significantly and carnitine slightly; coenzyme Q10 did not alter liver pathology; steroids did nothing. The authors state that controlled trials in children were not feasible, which is why the mouse was used; carnitine plus dextrose gave the best results. Mice.

86. 86
    Upadhyay SN, Kaushic C, Talwar GP. Antifertility effects of neem (Azadirachta indica) oil by single intrauterine administration: a novel method for contraception.Proc Biol Sci · 1990 · 242(1305):175-9doi:10.1098/rspb.1990.0121 · PMID 1983033

    A single 100 microl of neem oil into the rat uterus blocked fertility for 107 to 180 days, reversibly, with a pre-implantation block confined to the treated horn and leukocyte infiltration of the uterine lining; ovaries unaffected. Rats.

87. 87
    Upadhyay S, Dhawan S, Sharma MG, Talwar GP. Long-term contraceptive effects of intrauterine neem treatment (IUNT) in bonnet monkeys: an alternate to intrauterine contraceptive devices (IUCD).Contraception · 1994 · 49(2):161-9doi:10.1016/0010-7824(94)90091-4 · PMID 8143455

    One ml of neem oil by intrauterine catheter blocked fertility in bonnet monkeys for 7 to 12 months, reversibly, with more MHC class II antigen-presenting cells in the endometrium. Monkeys.

88. 88
    Garg S, Talwar GP, Upadhyay SN. Immunocontraceptive activity guided fractionation and characterization of active constituents of neem (Azadirachta indica) seed extracts.J Ethnopharmacol · 1998 · 60(3):235-46doi:10.1016/s0378-8741(97)00157-8 · PMID 9613837

    The intrauterine antifertility activity was traced by fractionation to a mixture of six saturated and unsaturated free fatty acids and their methyl esters, not to azadirachtin or any limonoid. Rats; a point about which molecules do what.

89. 89
    Talwar GP, Pal R, Singh O, et al. Safety of intrauterine administration of purified neem seed oil (Praneem Vilci) in women & effect of its co-administration with the heterospecies dimer birth control vaccine on antibody response to human chorionic gonadotropin.Indian J Med Res · 1995 · 102:66-70PMID 8834816

    Phase I: a single intrauterine instillation of purified neem oil in 18 women who had already had tubal sterilization; no untoward reaction, menstrual pattern and ovulation unchanged, one asymptomatic non-specific endometritis on biopsy, transient eosinophilia in two. Safety in sterilized women; it could not and did not test contraception.

90. 90
    Joshi SN, Katti U, Godbole S, et al. Phase I safety study of Praneem polyherbal vaginal tablet use among HIV-uninfected women in Pune, India.Trans R Soc Trop Med Hyg · 2005 · 99(10):769-74doi:10.1016/j.trstmh.2005.01.007 · PMID 16084547

    20 women using the tablet vaginally for 14 days: nine (45%) had 17 episodes of genital irritation, eight (40%) transient itching, two petechial hemorrhages on colposcopy that resolved; no serious events. Open label, 20 women; the abstract states the product 'also has contraceptive properties' on the strength of animal data.

91. 91
    Joshi SN, Dutta S, Kumar BK, et al. Expanded safety study of Praneem polyherbal vaginal tablet among HIV-uninfected women in Pune, India: a phase II clinical trial report.Sex Transm Infect · 2008 · 84(5):343-7doi:10.1136/sti.2007.029207 · PMID 18426844

    100 low-risk women randomized to Praneem or placebo before each act of sex for six months, 601 person-months: transient genital discomfort the main adverse event, no serious events, no HIV seroconversions, and the authors' warning that after nonoxynol-9, cellulose sulfate and Carraguard failed, 'additional preclinical evaluations should be undertaken urgently before moving into effectiveness studies'. A safety trial; no efficacy was tested and no efficacy trial followed in PubMed.

92. 92
    Rochanakij S, Thebtaranonth Y, Yenjai C, Yuthavong Y. Nimbolide, a constituent of Azadirachta indica, inhibits Plasmodium falciparum in culture.Southeast Asian J Trop Med Public Health · 1985 · 16(1):66-72PMID 3895455

    Nimbolide inhibited Plasmodium falciparum line K1 in culture with an EC50 of about 2.0 microM (0.95 microg/ml); the crude aqueous extract 115 microg/ml and the ethanol extract 5.0 microg/ml. Neither the extracts nor nimbolide had any activity against Plasmodium berghei in mice, by ingestion or by subcutaneous injection. Culture and mice; the first reversal of the antimalarial claim.

93. 93
    Khalid SA, Farouk A, Geary TG, Jensen JB. Potential antimalarial candidates from African plants: an in vitro approach using Plasmodium falciparum.J Ethnopharmacol · 1986 · 15(2):201-9doi:10.1016/0378-8741(86)90156-x · PMID 3520157

    Of 21 compounds from nine African medicinal plants, most were inactive; the limonoid gedunin had an IC50 of about 1 microM after 48 hours and 0.3 microM after 96 hours, 'roughly equivalent to quinine'. In vitro.

94. 94
    Khalid SA, Duddeck H, Gonzalez-Sierra M. Isolation and characterization of an antimalarial agent of the neem tree Azadirachta indica.J Nat Prod · 1989 · 52(5):922-6doi:10.1021/np50065a002 · PMID 2607354

    Isolation and structure elucidation by one- and two-dimensional nuclear magnetic resonance of gedunin as the antimalarial agent of neem. Chemistry; no in vivo work.

95. 95
    Fokou PVT, Tali MBT, Mbouna CDJ, et al. Natural products as transmission-blocking agents against malaria: a comprehensive review of bioactive compounds and their therapeutic potential.Malar J · 2025 · 24(1):164doi:10.1186/s12936-025-05395-6 · PMID 40420292

    2025 review naming Azadirachta indica among extracts with potent gametocytocidal effects and stating that 'despite promising in vitro and in vivo data, the transition of these compounds to clinical applications remains limited'. Review.

96. 96
    Pai MR, Acharya LD, Udupa N. The effect of two different dental gels and a mouthwash on plaque and gingival scores: a six-week clinical study.Int Dent J · 2004 · 54(4):219-23doi:10.1111/j.1875-595x.2004.tb00283.x · PMID 15335093

    Open-label six-week study in 48 subjects in four groups: a neem-extract gel reduced plaque and gingival scores against control and did not differ from a chlorhexidine gel. Small, open label; first author at a pharmaceutical research center (Zydus).

97. 97
    Nesari TM, Bhardwaj A, ShriKrishna R, et al. Neem (Azadirachta Indica A. Juss) Capsules for Prophylaxis of COVID-19 Infection: A Pilot, Double-Blind, Randomized Controlled Trial.Altern Ther Health Med · 2021 · 27(S1):196-203PMID 33891569

    190 contacts of COVID-19 patients randomized to 50 mg of a 'proprietary, patent-pending' neem leaf extract or placebo twice daily for 28 days: among 154 completing per protocol, 3 versus 8 infections, all asymptomatic, relative risk 0.45. Eleven events in total, a proprietary product, and a per-protocol analysis: a pilot, as titled.

98. 98
    Pingali U, Vuppalanchi B, Nutalapati C, Gundagani S. Aqueous Azadirachta indica (Neem) Extract Attenuates Insulin Resistance to Improve Glycemic Control and Endothelial Function in Subjects with Metabolic Syndrome.J Med Food · 2021 · 24(11):1135-1144doi:10.1089/jmf.2020.4838 · PMID 34582720

    Randomized placebo-controlled trial, 20 per group, of 125, 250 or 500 mg of an aqueous leaf-and-twig extract twice daily for 12 weeks in metabolic syndrome: dose-dependent improvement in fasting and post-meal glucose, insulin resistance, HbA1c, endothelial function and inflammatory markers; registered CTRI/2019/03/018034. A single-center trial of 80 people; the abstract gives no absolute numbers and the funding is not stated in it.

99. 99
    Widasmara D, Yuniaswan AP, Listy Pramita V. Comparative study of 5% permethrin, 10% neem leaf extract, and 10% brotowali stem extract for scabies treatment.Front Med (Lausanne) · 2025 · 12:1672149doi:10.3389/fmed.2025.1672149 · PMID 41488098

    61 male boarding-school students randomized to permethrin 5%, neem leaf extract 10% lotion or brotowali: lesion counts fell in all three with no difference in physician global assessment, quality of life improved earlier with the herbal lotions, no adverse effects; adherence, application time and reinfestation were not controlled. About 20 per arm; the first randomized neem scabies trial in people this page found, run in 2021 and published in December 2025.

100. 100
     Garg S, Taluja V, Upadhyay SN, Talwar GP. Studies on the contraceptive efficacy of Praneem polyherbal cream.Contraception · 1993 · 48(6):591-6doi:10.1016/0010-7824(93)90120-v · PMID 8131399

     Praneem polyherbal cream (purified neem seed extract, Sapindus saponins and quinine hydrochloride) was non-irritant in rabbits and 21-day human skin tests and 'has shown high contraceptive efficacy in rabbits and in monkeys' after vaginal application. Animals; developed at India's National Institute of Immunology.

101. 101
     Mukherjee S, Lohiya NK, Pal R, Sharma MG, Talwar GP. Purified neem (Azadirachta indica) seed extracts (Praneem) abrogate pregnancy in primates.Contraception · 1996 · 53(6):375-8doi:10.1016/0010-7824(96)00088-1 · PMID 8773426

     Oral Praneem for six days from day 25 of the cycle terminated confirmed early pregnancy in baboons and bonnet monkeys, with falling gonadotropin and progesterone and return of normal cycles; peanut oil controls stayed pregnant. Primates; an abortifacient effect, which matters for anyone pregnant who takes neem.

102. 102
     Moore SJ, Lenglet A, Hill N. Field evaluation of three plant-based insect repellents against malaria vectors in Vaca Diez Province, the Bolivian Amazon.J Am Mosq Control Assoc · 2002 · 18(2):107-10PMID 12083351

     Human landing catches against Anopheles darlingi: a eucalyptus product gave 96.89% protection for 4 h and deet 84.81%; the neem-based product 'did not provide significant protection from mosquito bites'. A field negative against the 1993 claim.

103. 103
     Sharma VP, Ansari MA, Razdan RK. Mosquito repellent action of neem (Azadirachta indica) oil.J Am Mosq Control Assoc · 1993 · 9(3):359-60PMID 8245950

     2% neem oil in coconut oil on volunteers' skin gave 'complete protection for 12 h' from all anopheline bites. A short note from India's Malaria Research Centre.

104. 104
     Sharma RS, Mathur AK, Chandhiok N, et al. Phase II clinical trial with Praneem polyherbal tablets for assessment of their efficacy in symptomatic women with abnormal vaginal discharge (an ICMR task force study).Trans R Soc Trop Med Hyg · 2009 · 103(2):167-72doi:10.1016/j.trstmh.2008.09.012 · PMID 18990416

     141 women at six Indian Council of Medical Research centers used the tablet for seven days: 97% reported complete (44%) or partial (53%) relief, 74% cured on speculum examination, microscopic cure in all Trichomonas, 77% of Candida and 68% of bacterial vaginosis cases; 55% had transient burning. No control group.

105. 105
     Salhan S, Tripathi V, Sehgal R, Kumar G, Talwar GP, Chatterjee A. A phase II randomized controlled trial to evaluate the safety and efficacy of Praneem polyherbal vaginal tablets compared with betadine vaginal pessary in women with symptoms of abnormal vaginal discharge.Asia Pac J Public Health · 2009 · 21(4):461-8doi:10.1177/1010539509344610 · PMID 19783561

     99 women randomized to Praneem or povidone-iodine pessary: symptom relief in 92% versus 81.6%; the authors call for a phase III trial 'for definitive conclusions'. Symptom-based; the developer (Talwar) is an author.

106. 106
     Patel Y, Gopalan S, Bagga R, Sharma M, Chopra S, Sethi S. A randomized trial comparing a polyherbal pessary (a complementary and alternative medicine) with Ginlac-V pessary (containing clotrimazole, tinidazole and lactobacilli) for treatment of women with symptomatic vaginal discharge.Arch Gynecol Obstet · 2008 · 278(4):341-7doi:10.1007/s00404-008-0568-9 · PMID 18236058

     100 women, 50 per arm, in Chandigarh: symptomatic relief 78% with Praneem and 86% with the antifungal-antiprotozoal pessary; only 36% had laboratory-confirmed infection; neither drug was efficacious in candidiasis; Ginlac-V cured bacterial vaginosis in 100% and Praneem showed a non-significant trend; vaginal irritation 16% with Praneem versus 4%. An independent trial with a harder look.

107. 107
     Sharma S, Saimbi CS, Koirala B, Shukla R. Effect of various mouthwashes on the levels of interleukin-2 and interferon-gamma in chronic gingivitis.J Clin Pediatr Dent · 2008 · 32(2):111-4doi:10.17796/jcpd.32.2.u01p135561161476 · PMID 18389675

     Double-blind, 80 adolescents (mean age 16) randomized to neem, essential-oil, povidone-iodine or chlorhexidine mouthwash for two weeks: the other three significantly lowered plaque, gingival index and the two gingival cytokines; neem lowered them 'to a certain level' that was not statistically significant. A negative trial for neem, 20 per arm.

108. 108
     Dallaqua B, Saito FH, Rodrigues T, et al. Treatment with Azadirachta indica in diabetic pregnant rats: negative effects on maternal outcome.J Ethnopharmacol · 2012 · 143(3):805-11doi:10.1016/j.jep.2012.07.023 · PMID 22921950

     In non-diabetic and mildly diabetic pregnant rats, neither neem oil nor azadirachtin lowered glucose, both raised the share of fetuses small for gestational age, and both raised lipid peroxidation. Rats; a negative result for the diabetes claim under the one condition tested.

109. 109
     Semmler M, Abdel-Ghaffar F, Gestmann F, et al. Randomized, investigator-blinded, controlled clinical study with lice shampoo (Licener) versus dimethicone (Jacutin Pedicul Fluid) for the treatment of infestations with head lice.Parasitol Res · 2017 · 116(7):1863-1870doi:10.1007/s00436-017-5461-7 · PMID 28488042

     119 Egyptian children: cure after one application 60/60 (100%, 95% CI 94.04-100) with the neem shampoo against 54/57 (94.74%, 85.38-98.90) with dimethicone, p = 0.112; combined success 100% versus 90.7%, p = 0.024. The trial's stated main objective was a cure rate of at least 85% after a single application, which both products met; it was not powered to show superiority. One author is from the manufacturer, Alpha-Biocare, who is also an author of an earlier uncontrolled series of 65 patients with the same product [174]; both products are physical pediculicides.

110. 110
     Al-Quraishy S, Abdel-Ghaffar F, Mehlhorn H. Head louse control by suffocation due to blocking their oxygen uptake.Parasitol Res · 2015 · 114(8):3105-10doi:10.1007/s00436-015-4528-6 · PMID 25990060

     Microscopy showing the neem-seed shampoo (Licener) filling the louse tracheal system; death of all stages within 3-10 minutes. The mode of action is physical, not azadirachtin's hormonal one.

111. 111
     Planas J, Moraleda-Berral P, Checa R, et al. Efficacy of neem oil (Azadirachta indica) as an antiparasitic treatment for scorpions (Androctonus bicolor) and its impact on venom production.Vet Parasitol · 2026 · 344:110748doi:10.1016/j.vetpar.2026.110748 · PMID 41903410

     A 1% neem oil spray every 72 h for five sessions cleared Pimeliaphilus mites from 40 captive scorpions with no mortality or cuticle damage; venom yield rose. Included as the newest and oddest animal use; an uncontrolled before-after study.

112. 112
     Andrade Coelho CA, de Souza NA, Feder MD, et al. Effects of azadirachtin on the development and mortality of Lutzomyia longipalpis larvae (Diptera: Psychodidae: Phlebotominae).J Med Entomol · 2006 · 43(2):262-6doi:10.1603/0022-2585(2006)043[0262:eoaotd]2.0.co;2 · PMID 16619609

     Azadirachtin A in the diet at 0.1-10 microg/mg blocked molting of sand-fly larvae and raised mortality; adding ecdysone (1 microg/mg) to the same diet counteracted both. Laboratory insects; the rescue experiment.

113. 113
     Suman DS, Wang Y, Bilgrami AL, Gaugler R. Ovicidal activity of three insect growth regulators against Aedes and Culex mosquitoes.Acta Trop · 2013 · 128(1):103-9doi:10.1016/j.actatropica.2013.06.025 · PMID 23860181

     Freshly laid Aedes albopictus eggs: 42.9% hatching inhibition at 1 ppm azadirachtin against 80.6% for pyriproxyfen; embryonated eggs resistant. The abstract files azadirachtin as an 'ecdysone agonist', the opposite of the usual description, which this page flags. Laboratory mosquito eggs.

114. 114
     Warbrick EV, Barker GC, Rees HH, Howells RE. The effect of invertebrate hormones and potential hormone inhibitors on the third larval moult of the filarial nematode, Dirofilaria immitis, in vitro.Parasitology · 1993 · 107(Pt 4):459-63doi:10.1017/s0031182000067822 · PMID 8278225

     Azadirachtin at 10^-5 M prevented most heartworm larvae from molting to the fourth stage in culture, while 20-hydroxyecdysone brought the molt forward. In vitro, on a dog parasite; the only nematode-molting datum on this page.

115. 115
     Nehra R, Dhanda S, Singh H, Bharti PK, Singh K. Biopesticides as an eco-friendly approach to mosquitoes and their role in reducing transmission of vector-borne disease: updates and prospects.Pest Manag Sci · 2026 · 82(10):9387-9400doi:10.1002/ps.71077 · PMID 42426957

     2026 review from India's National Institute of Malaria Research describing azadirachtin as inhibiting ecdysone signaling and reducing mosquito fecundity, alongside bacterial and fungal agents. Review.

116. 116
     Scudeler EL, Garcia AS, Padovani CR, Pinheiro PF, dos Santos DC. Cytotoxic effects of neem oil in the midgut of the predator Ceraeochrysa claveri.Micron · 2015 · 80:96-111doi:10.1016/j.micron.2015.10.005 · PMID 26520254

     Lacewing larvae fed prey eggs treated with 0.5-2% neem oil became adults with lysed midgut cells, swollen mitochondria and a stress response that did not repair all the damage; a sub-lethal effect on a beneficial insect regarded as 'safe' for it. Insect histology.

117. 117
     Bernardi MM, Dias SG, Barbosa VE. Neurotoxicity of neem commercial formulation (Azadirachta indica A. Juss) in adult zebrafish (Danio rerio).Environ Toxicol Pharmacol · 2013 · 36(3):1276-82doi:10.1016/j.etap.2013.10.002 · PMID 24211596

     Adult zebrafish in 20 or 40 microl/l of a commercial neem formulation showed reduced activity, tremors at the higher dose and anxiety-like behavior, with learning and memory unaffected. Fish behavior.

118. 118
     Sacco D, Velíšek J, Mikušková N. Toxicological effects of azadirachtin on aquatic species: A review of its role in biopesticides.Aquat Toxicol · 2025 · 288:107547doi:10.1016/j.aquatox.2025.107547 · PMID 40897009

     2025 review: azadirachtin degrades fast in soil but has lethal and sub-lethal effects on fish and aquatic invertebrates, with unknowns about its metabolites and formulations. Environmental, not human, toxicity.

119. 119
     Ferreira MB, Costa GAC, Dos Santos PC, et al. Impacts of encapsulated azadirachtin (Azadirachta indica) in the diet of Nile tilapia (Oreochromis niloticus).Fish Shellfish Immunol · 2026 · 171:111156doi:10.1016/j.fsi.2026.111156 · PMID 41611136

     400 juvenile tilapia fed 0, 20, 70, 120 or 170 mg azadirachtin per kg of feed for 45 days: growth unchanged; at 170 mg/kg the blood plasma was more active against Aeromonas hydrophila and intestinal streptococci were fewer, but the structure of kidney and liver was damaged, malondialdehyde rose in liver tissue at 120 mg/kg, and the authors advise caution about putting azadirachtin in fish feed. Fish.

120. 120
     Shahzad M, Ahmad HA, Ambadi S, Peterson S, Yang I. Traditional Oral Hygiene Practices and Their Effectiveness: A Systematic Review of the Evidence.Oral Health Prev Dent · 2026 · 24:107-118doi:10.3290/j.ohpd.c_2475 · PMID 41670379

     2026 systematic review of 31 studies to April 2024: miswak, the arak-tree chewing stick, reduced plaque and gingival inflammation comparably to or better than toothbrushes, and herbal toothpastes and rinses containing neem, clove, turmeric and ginger were as effective as fluoridated products in the short term, the neem evidence in it being for those products rather than for neem chewing sticks, with 'methodological limitations and variability' requiring 'cautious interpretation'. Systematic review.

121. 121
     Chatterjee A, Saluja M, Singh N, Kandwal A. To evaluate the antigingivitis and antiplaque effect of an Azadirachta indica (neem) mouthrinse on plaque induced gingivitis: A double-blind, randomized, controlled trial.J Indian Soc Periodontol · 2011 · 15(4):398-401doi:10.4103/0972-124X.92578 · PMID 22368367

     45 subjects in three groups of 15 (neem, chlorhexidine, saline) for a month: neem and chlorhexidine both reduced gingival, bleeding and plaque indices against saline over 21 days, and were 'equally effective'. Fifteen per arm; equivalence was not formally tested.

122. 122
     Balappanavar AY, Sardana V, Singh M. Comparison of the effectiveness of 0.5% tea, 2% neem and 0.2% chlorhexidine mouthwashes on oral health: a randomized control trial.Indian J Dent Res · 2013 · 24(1):26-34doi:10.4103/0970-9290.114933 · PMID 23852229

     30 volunteers aged 18-25, ten per arm, three weeks: all three reduced plaque and gingival scores; tea came out best for plaque and neem and tea better than chlorhexidine for gingiva. Ten per arm.

123. 123
     Jalaluddin M, Rajasekaran UB, Paul S, Dhanya RS, Sudeep CB, Adarsh VJ. Comparative Evaluation of Neem Mouthwash on Plaque and Gingivitis: A Double-blind Crossover Study.J Contemp Dent Pract · 2017 · 18(7):567-571doi:10.5005/jp-journals-10024-2085 · PMID 28713109

     Double-blind crossover in 40 adults, 2% neem against 0.2% chlorhexidine with a one-week washout: slight plaque reductions in both phases with no significant difference between groups. Forty subjects.

124. 124
     Nishad A, Sreesan NS, Joy J, Lakshmanan L, Thomas J, Anjali VA. Impact of Mouthwashes on Antibacterial Activity of Subjects with Fixed Orthodontic Appliances: A Randomized Clinical Trial.J Contemp Dent Pract · 2017 · 18(12):1112-1116doi:10.5005/jp-journals-10024-2185 · PMID 29208784

     60 orthodontic patients, 20 per arm, 30 days: chlorhexidine and neem mouthwashes both reduced plaque and gingival indices and S. mutans counts against water. Twenty per arm.

125. 125
     Tidke S, Chhabra GK, Madhu PP, Reche A, Wazurkar S, Singi SR. The Effectiveness of Herbal Versus Non-Herbal Mouthwash for Periodontal Health: A Literature Review.Cureus · 2022 · 14(8):e27956doi:10.7759/cureus.27956 · PMID 36120261

     Review of randomized trials 2001-2021 (seven included): herbal mouthwashes as effective as non-herbal in the short term, 'however, the evidence is based on low-quality trials'. Narrative review.

126. 126
     Abdulameer H, Badeia R. Comparative evaluation of two herbal mouthwashes and chlorhexidine on calprotectin in gingivitis patients: a double-blind randomized controlled trial.Georgian Med News · 2026 · (376-377):331-340PMID 42830072

     44 men with gingivitis in four groups of 11 for seven days; a propolis-neem-chlorophyll rinse, a propolis-green tea rinse, chlorhexidine and water all improved clinical indices and lowered gingival-fluid calprotectin with no difference between groups (p = 0.760). Eleven per arm, seven days, and water did as well as everything else.

127. 127
     Nandlal B, Sreenivasan PK, Shashikumar P, Devishree G, Bettahalli Shivamallu A. A randomized clinical study to examine the oral hygiene efficacy of a novel herbal toothpaste with zinc over a 6-month period.Int J Dent Hyg · 2021 · 19(4):440-449doi:10.1111/idh.12505 · PMID 33866666

     180 enrolled, 150 analyzed, six months: a toothpaste with six herbal ingredients including neem plus zinc and fluoride reduced plaque, gingival and bleeding scores 23.5%, 25.6% and 73.3% more than fluoride paste. Neem is one of seven actives, so nothing can be attributed to it; the companion paper names the manufacturer as Colgate-Palmolive India, a commercial interest.

128. 128
     Sreenivasan PK, Kakarla VVP, Sharda S, Setty Y. The effects of a novel herbal toothpaste on salivary lactate dehydrogenase as a measure of cellular integrity.Clin Oral Investig · 2021 · 25(5):3021-3030doi:10.1007/s00784-020-03623-8 · PMID 33064207

     Double-blind 12-week study of 70 adults of the same herbal-zinc toothpaste (identified as Ved Shakti, Colgate Palmolive India) against Colgate Cavity Protection: 9.5-15.4% lower salivary lactate dehydrogenase and 6.4-16.2% less plaque. The product and comparator are both the sponsor's.

129. 129
     Hosny NS, El Khodary SA, El Boghdadi RM, Shaker OG. Effect of Neem (Azadirachta indica) versus 2.5% sodium hypochlorite as root canal irrigants on the intensity of post-operative pain and the amount of endotoxins in mandibular molars with necrotic pulps: a randomized controlled trial.Int Endod J · 2021 · 54(9):1434-1447doi:10.1111/iej.13532 · PMID 33884661

     50 patients, 25 per arm, Cairo: pain scores lower with neem only at 24 h after instrumentation; both irrigants cut endotoxin (8% hypochlorite, 18% neem) and neither cleared it. Small, with one significant time point among eight.

130. 130
     Tanvir Z, Jabin Z, Agarwal N, Anand A, Waikhom N. Comparative evaluation of antimicrobial efficacy of nanosilver solution, Azadirachta indica, sodium hypochlorite, and normal saline as root canal irrigants in primary teeth.J Indian Soc Pedod Prev Dent · 2023 · 41(1):76-82doi:10.4103/jisppd.jisppd_74_23 · PMID 37282416

     80 children, 20 per arm: nanosilver reduced anaerobic colony counts most, then 2.5% hypochlorite, then neem. A ranking in which neem came last of the three actives.

131. 131
     Kendre SB, Bhatane AU, Dadpe MV, Kale YJ, Dahake PT. Comparative evaluation of antibacterial efficacy of sequential herbal irrigation with conventional irrigation in endodontic therapy of primary teeth: A randomized controlled trial.J Indian Soc Pedod Prev Dent · 2025 · 43(3):410-417doi:10.4103/jisppd.jisppd_253_25 · PMID 41026567

     Split-mouth trial in 15 children: a sequence of miswak, 20% neem, tea tree oil and phytic acid reduced Enterococcus faecalis 92.2% against 90.54% for hypochlorite-EDTA-chlorhexidine (p = 0.025), Prevotella no different. Fifteen children; neem is one of four agents.

132. 132
     Susila AV, Sai S, Sharma N, Balasubramaniam A, Veronica AK, Nivedhitha S. Can natural irrigants replace sodium hypochlorite? A systematic review.Clin Oral Investig · 2023 · 27(5):1831-1849doi:10.1007/s00784-023-04913-7 · PMID 36808559

     Ten clinical studies in about 442 patients, too heterogeneous to pool. Antimicrobial efficacy was similar for several natural irrigants and sodium hypochlorite; neem and some neem combinations were superior in individual comparisons and post-operative pain was less with neem; the review's conclusion is that the natural irrigants 'are not more efficacious than NaOCl' and 'cannot replace NaOCl routinely'.

133. 133
     Kishan KV, Shah N, Mamatha KV, Sreekumari L, Parikh M. A randomized clinical comparative evaluation of interappointment flare-ups on placing neem, garlic, Triphala, and calcium hydroxide as intracanal medicament, in primary endodontic lesion.J Conserv Dent Endod · 2025 · 28(2):155-160doi:10.4103/JCDE.JCDE_824_24 · PMID 40046471

     240 patients, 30 per subgroup, pain and flare-ups scored at days 1, 3, 7 and 14 after a dressing was left in the canal: Triphala came out as a possible alternative to calcium hydroxide while 'neem and garlic showed poor results'. A negative randomized result for neem in this role.

134. 134
     Vashisht R, Kumar U, Kakkar AK, Katare OP, Patil UK, Brar PK. Post-operative pain in molar teeth with symptomatic apical periodontitis using cryotreated neem extract and cryotreated 2.5% sodium hypochlorite: A randomised double-blinded clinical trial.Bioinformation · 2026 · 22(4):2701-2705doi:10.6026/973206300222701 · PMID 42282414

     108 patients in four groups, 27 per arm: chilled irrigants gave lower pain scores at 24 and 48 hours than room-temperature irrigants, and chilled neem extract was comparable to chilled hypochlorite. Pain only, one week of follow-up.

135. 135
     Mehta V, Gaur A, Chaurasia M, et al. Efficacy of Herbal Fluoride Toothpaste in Preventing Early Childhood Caries.J Pharm Bioallied Sci · 2025 · 17(Suppl 4):S3398-S3400doi:10.4103/jpbs.jpbs_1437_25 · PMID 41522918

     120 children aged 3-5 randomized to a toothpaste containing neem, clove and 500 ppm fluoride or a plain 500 ppm fluoride paste: mean increase in decayed, missing or filled baby teeth over 12 months 0.42 against 0.95 (p = 0.02), new lesions in 16.6% against 28.3%. The only trial on this page that counted cavities; 60 per arm, three active ingredients, and a conference-style supplement issue.

136. 136
     Oak A, Bachubhai Sapariya D, Nayak C, Kumar Reddy AS, Lakshmi RS, Dalal D. A Randomized Double-Blind Clinical Trial Evaluates the Efficacy of Alternative Herbal Mouthwashes.Cureus · 2023 · 15(6):e40394doi:10.7759/cureus.40394 · PMID 37456435

     125 patients in five groups including a neem-containing herbal rinse and chlorhexidine: no significant difference among the three herbal rinses; the abstract does not report the herbal-versus-chlorhexidine comparison. Twenty-five per arm.

137. 137
     Charles V, Charles SX. The use and efficacy of Azadirachta indica ADR ('Neem') and Curcuma longa ('Turmeric') in scabies. A pilot study.Trop Geogr Med · 1992 · 44(1-2):178-81PMID 1496714

     A neem-turmeric paste was used on 814 people with scabies in Tamil Nadu; cure within 3 to 15 days in 97%; no adverse reactions noticed. No control group, no blinding, no mite confirmation stated; the authors themselves say further research is needed.

138. 138
     Deng Y, Shi D, Yin Z, et al. Acaricidal activity of petroleum ether extract of neem (Azadirachta indica) oil and its four fractions separated by column chromatography against Sarcoptes scabiei var. cuniculi larvae in vitro.Exp Parasitol · 2012 · 130(4):475-7doi:10.1016/j.exppara.2012.02.007 · PMID 22349080

     Rabbit scabies mites dipped in a petroleum-ether extract of neem oil: LC50 70.9 ml/l at 24 h, with two of four chromatographic fractions active. In vitro on an animal mite.

139. 139
     Abdel-Ghaffar F, Semmler M. Efficacy of neem seed extract shampoo on head lice of naturally infected humans in Egypt.Parasitol Res · 2006 · 100(2):329-32doi:10.1007/s00436-006-0264-2 · PMID 16900389

     60 heavily infested Egyptian children aged 4-15: a neem seed extract shampoo left on 10 to 30 minutes killed all louse stages with no irritation. Uncontrolled.

140. 140
     Mehlhorn H, Abdel-Ghaffar F, Al-Rasheid KA, Schmidt J, Semmler M. Ovicidal effects of a neem seed extract preparation on eggs of body and head lice.Parasitol Res · 2011 · 109(5):1299-302doi:10.1007/s00436-011-2374-8 · PMID 21484346

     Five minutes in the shampoo stopped all hatching of head and body louse eggs (controls 76% and 93%); the authors attribute it to blocking of the eggs' air pores. Laboratory; product named as Wash Away Louse, from a German company.

141. 141
     Abdel-Ghaffar F, Abdel-Aty M, Rizk I, Al-Quraishy S, Semmler M, Gestmann F, Hoff NP. Head lice in progress: what could/should be done-a report on an in vivo and in vitro field study.Parasitol Res · 2016 · 115(11):4245-4249doi:10.1007/s00436-016-5202-3 · PMID 27481171

     65 patients with head lice treated with the neem-seed shampoo in individual therapy trials, with no relapses among those successfully treated. Uncontrolled, and an author is from the manufacturer, Alpha-Biocare, the same commercial interest as in the 2017 randomized trial [64].

142. 142
     Elson L, Randu K, Feldmeier H, Fillinger U. Efficacy of a mixture of neem seed oil (Azadirachta indica) and coconut oil (Cocos nucifera) for topical treatment of tungiasis. A randomized controlled, proof-of-principle study.PLoS Negl Trop Dis · 2019 · 13(11):e0007822doi:10.1371/journal.pntd.0007822 · PMID 31756189

     96 Kenyan children with embedded sand fleas randomized to 20% neem oil in coconut oil or potassium permanganate foot baths: neem killed about 40% of fleas by day 6, no better than permanganate, but fleas aged faster (OR 3.4, 95% CI 1.22-9.49) and children were more often pain free (OR 3.5, p = 0.001). Registered trial; secondary outcomes only.

143. 143
     Farahnik B, Sharma D, Alban J, Sivamani R. Oral (Systemic) Botanical Agents for the Treatment of Psoriasis: A Review.J Altern Complement Med · 2017 · 23(6):418-425doi:10.1089/acm.2016.0324 · PMID 28157393

     Twelve controlled and uncontrolled trials of oral botanicals in psoriasis: 'while overall evidence is limited in quantity and quality', HESA-A, curcumin and neem extract seemed most efficacious. Review; one author works for a dermatology company (Dermveda).

144. 144
     Talbott W, Duffy N. Complementary and alternative medicine for psoriasis: what the dermatologist needs to know.Am J Clin Dermatol · 2015 · 16(3):147-65doi:10.1007/s40257-015-0128-6 · PMID 25904522

     Review placing neem among herbals with 'a smaller amount of evidence' than Mahonia and indigo naturalis, and warning of systemic toxicity and contact dermatitis from herbal supplements. Review.

145. 145
     Sabiha K, Misbahuddin M, Chowdhury ASM. Comparison between Combination of Neem, Propylene Glycol and Salicylic Acid with Salicylic Acid Alone Topically in Arsenical Palmar Keratosis.Mymensingh Med J · 2025 · 34(3):728-737PMID 40583662

     30 patients with arsenical keratosis of the palms in Bangladesh: nodule score fell from 19.6 to 8.1 with a neem-propylene glycol-salicylic acid lotion over 12 weeks, with a higher patient-rated improvement than salicylic acid alone. Fifteen per arm; three ingredients differ between arms.

146. 146
     Nelramachandrakumar D, Hariharan H, Nagamanickam S, Chander J SJU. Drug-aggravated Bullous Pemphigoid in a 47-year-old Asian Woman - A Case Report.Curr Drug Saf · 2023 · 18(3):379-382doi:10.2174/1574886317666220518091554 · PMID 35593331

     A woman with blisters on the lower limbs self-treated them with neem oil, after which blisters appeared over the body; biopsy confirmed bullous pemphigoid, an autoimmune blistering disease, and the authors present the case to warn against folk-medicine use in such a disease. One case, and the causal reading is the authors'.

147. 147
     Patel N, Knabel M, Speiser J. Botanical Briefs: Neem Oil (Azadirachta indica).Cutis · 2024 · 113(1):22-24doi:10.12788/cutis.0928 · PMID 38478934

     A short review with a case: a 77-year-old man with a hypopigmented rash on the lower back, flanks and buttocks after six months of repeated neem oil applied for arthritis and back pain. One case.

148. 148
     Vatandoost H, Hanafi-Bojd AA. Laboratory evaluation of 3 repellents against Anopheles stephensi in the Islamic Republic of Iran.East Mediterr Health J · 2008 · 14(2):260-7PMID 18561716

     On rabbit skin, the dose giving 50% repellency against Anopheles stephensi was 0.007 mg/cm2 for permethrin, 0.005 for deet and 0.191 for neem extract; neem 'was the least effective agent'. Laboratory.

149. 149
     Kebede Y, Gebre-Michael T, Balkew M. Laboratory and field evaluation of neem (Azadirachta indica A. Juss) and Chinaberry (Melia azedarach L.) oils as repellents against Phlebotomus orientalis and P. bergeroti (Diptera: Psychodidae) in Ethiopia.Acta Trop · 2010 · 113(2):145-50doi:10.1016/j.actatropica.2009.10.009 · PMID 19854142

     Against Phlebotomus orientalis, the sand fly that carries visceral leishmaniasis, 2% neem oil in coconut oil gave 96.28% protection for a mean of 7 hours 20 minutes and 5% gave 98.26% for 9 hours in the laboratory, with similar protection in field tests; chinaberry oil was comparable. Laboratory and field; the positive counterpart to the failed mosquito-repellent claim.

150. 150
     Faridah L, Nurhakim RA, Astuti RRUNW, et al. Effectiveness of mosquito repellents for human: A systematic review of human-based studies.Acta Trop · 2026 · 280:108182doi:10.1016/j.actatropica.2026.108182 · PMID 42285396

     Systematic review of 19 human studies published between 2015 and 2025: DEET lotions and ointments gave the longest complete protection times, often five to nine hours, while plant-based repellents including neem commonly gave two to four hours, improved in some microencapsulated or nanoemulsion formulations.

151. 151
     Korsah S, Nortey NND, Akakpo C, et al. Synergistic Combination of Cryptolepis sanguinolenta and Azadirachta indica Against 3D7 and DD2 Strains of Plasmodium falciparum.Adv Pharmacol Pharm Sci · 2026 · 2026:1273876doi:10.1155/adpp/1273876 · PMID 42677102

     2026, Ghana: aqueous extracts of both plants gave IC50 at or below 10 microg/ml against sensitive and resistant strains and were synergistic (combination index 0.15-0.52). In vitro only.

152. 152
     Happi GM, Nangmo PK, Dzouemo LC, Kache SF, Kouam ADK, Wansi JD. Contribution of Meliaceous plants in furnishing lead compounds for antiplasmodial and insecticidal drug development.J Ethnopharmacol · 2022 · 285:114906doi:10.1016/j.jep.2021.114906 · PMID 34910951

     Review of 124 Meliaceae compounds tested against 11 P. falciparum strains: limonoids were 51.6% of them and the most active, gedunin equal to quinine; azadirachtin gave near-100% larval mortality of Anopheles stephensi at 1 ppm; the authors call for ADMET studies, meaning absorption, distribution, metabolism, excretion and toxicity, before any drug. Review.

153. 153
     Tajbakhsh E, Kwenti TE, Kheyri P, Nezaratizade S, Lindsay DS, Khamesipour F. Antiplasmodial, antimalarial activities and toxicity of African medicinal plants: a systematic review of literature.Malar J · 2021 · 20(1):349doi:10.1186/s12936-021-03866-0 · PMID 34433465

     Systematic review of 200 papers covering 722 experiments on 502 plant species, 81.9% of them in vitro. Azadirachta indica was the most frequently investigated species, the species for which toxicity was most frequently reported, and among the species most frequently reported inactive; only one clinical trial of any African antimalarial plant had been conducted, and not of neem.

154. 154
     Bedri S, Khalil EA, Khalid SA, et al. Azadirachta indica ethanolic extract protects neurons from apoptosis and mitigates brain swelling in experimental cerebral malaria.Malar J · 2013 · 12:298doi:10.1186/1475-2875-12-298 · PMID 23984986

     In mice with Plasmodium berghei cerebral malaria, intraperitoneal ethanolic neem extract at 300-1,000 mg/kg reduced neuroinflammation, brain edema and neuronal apoptosis comparably to chloroquine and artemether. Mice; a histological, not a survival, endpoint.

155. 155
     Ali W, Ali M, Sehar M, et al. Antidiabetic Effect of Coronopus Didymus, Azadirachta Indica and Their Active Compound Quercetin Against STZ-NA Induced Diabetes Mice Models.Appl Biochem Biotechnol · 2026 · 198(11):8482-8507doi:10.1007/s12010-026-05851-2 · PMID 42496827

     2026: in streptozotocin-nicotinamide diabetic mice, neem extract alone lowered blood glucose and a three-way combination protected the liver. Mice; the current state of the diabetes claim.

156. 156
     Bandyopadhyay U, Biswas K, Sengupta A, et al. Clinical studies on the effect of Neem (Azadirachta indica) bark extract on gastric secretion and gastroduodenal ulcer.Life Sci · 2004 · 75(24):2867-78doi:10.1016/j.lfs.2004.04.050 · PMID 15454339

     Patients with acid problems given 30 mg of lyophilized bark extract twice daily for 10 days had 77% less gastric acid secretion; 30-60 mg twice daily for 10 weeks 'almost completely healed' duodenal ulcers on barium meal or endoscopy, with one esophageal and one gastric ulcer healed; blood tests normal. Uncontrolled, numbers not given in the abstract.

157. 157
     Jin M, Mengfan LV, Yu H, et al. Neem leaf extract alleviates LPS/D-GalN induced acute hepatitis in mice through its anti-inflammatory effects and activation of autophagy.Mol Immunol · 2025 · 180:33-43doi:10.1016/j.molimm.2025.02.015 · PMID 39999524

     2025: an ethanolic leaf extract containing azadirachtin and nimbolide reduced mortality and transaminases in a mouse model of acute hepatitis, acting through NF-kappaB and, partly, autophagy. Mice.

158. 158
     Bedi O, Bijjem KRV, Kumar P, Gauttam V. Herbal Induced Hepatoprotection and Hepatotoxicity: A Critical Review.Indian J Physiol Pharmacol · 2016 · 60(1):6-21PMID 29953177

     Review of herbal liver effects that lists Azadirachta indica among plants 'found to be hepatotoxic' alongside Lantana camara, comfrey and the death-cap mushroom, while listing other plants as hepatoprotective. Review; the neem entry rests on the oil poisonings.

159. 159
     Burton MJ, Clarkson JE, Goulao B, et al. Antimicrobial mouthwashes (gargling) and nasal sprays administered to patients with suspected or confirmed COVID-19 infection to improve patient outcomes and to protect healthcare workers treating them.Cochrane Database Syst Rev · 2020 · 9(9):CD013627doi:10.1002/14651858.CD013627.pub2 · PMID 32936948

     Cochrane review to June 2020 found no completed studies; among 16 ongoing trials were a neem-oil-containing spray and a neem extract gargle. Included to date the claim, not to support it.

160. 160
     Kaushic C, Upadhyay S. Mode of long-term antifertility effect of intrauterine neem treatment (IUNT).Contraception · 1995 · 51(3):203-7doi:10.1016/0010-7824(95)00018-6 · PMID 7621690

     In rats the effect was not uterine unresponsiveness to hormones but degeneration of embryos on the treated side between days 3 and 5 after mating. Rats.

161. 161
     Garg S, Doncel G, Chabra S, Upadhyay SN, Talwar GP. Synergistic spermicidal activity of neem seed extract, reetha saponins and quinine hydrochloride.Contraception · 1994 · 50(2):185-90doi:10.1016/0010-7824(94)90054-x · PMID 7956217

     Sander-Cramer test on human sperm: purified neem seed extract immobilized 100% of sperm within 20 seconds at 25%, reetha saponins at 0.05%, quinine at 0.346%, with synergy in combination. In vitro on human semen.

162. 162
     Joglekar NS, Joshi SN, Deshpande SS, Parkhe AN, Katti UR, Mehendale SM. Acceptability and adherence: findings from a Phase II study of a candidate vaginal microbicide, 'Praneem polyherbal tablet', in Pune, India.Trans R Soc Trop Med Hyg · 2010 · 104(6):412-5doi:10.1016/j.trstmh.2009.12.007 · PMID 20096909

     Of the 100 phase II participants, 70.5% reported full adherence; odor was a concern in both arms and women on Praneem were more likely to be partially adherent (adjusted OR 3.8, 95% CI 1.4-10.8) because of the experience during sex. Acceptability study.

163. 163
     Katafuchi Y, Yoshida I, Yamashita F, Sinniah D. Effects of drugs and fatty acids related to Reye syndrome on brain water content in rats.Acta Paediatr Jpn · 1989 · 31(2):115-9doi:10.1111/j.1442-200x.1989.tb01276.x · PMID 2516690

     Intraperitoneal margosa oil raised rat brain water content and, like valproate and 4-pentenoic acid, caused seizures in every animal. Rats; the one model in which the oil did produce brain swelling.

164. 164
     Tiwary BS. Neem leaf poisoning.J Assoc Physicians India · 1985 · 33(12):817PMID 3837008

     A one-page letter, no abstract, indexed under plant extracts and terpenes: the only PubMed-indexed report this page found under the heading of neem leaf poisoning in adults.

165. 165
     Durasnel P, Vanhuffel L, Blonde R, et al. [Severe poisoning by plants used for traditional medicine in Mayotte].Bull Soc Pathol Exot · 2014 · 107(5):306-11doi:10.1007/s13149-014-0400-7 · PMID 25301110

     Three adults severely poisoned by plants used in traditional treatment in Mayotte; the neurological signs and multiorgan failure are attributed to Azadirachta indica or Melia azedarach, the cardiac presentations to other plants. All three recovered. In French with an English abstract; the authors emphasize that identifying the plant and proving its responsibility is rarely possible.

166. 166
     Ibrahim IA, Khalid SA, Omer SA, Adam SE. On the toxicology of Azadirachta indica leaves.J Ethnopharmacol · 1992 · 35(3):267-73doi:10.1016/0378-8741(92)90024-l · PMID 1548899

     Chicks fed diets containing 2% or 5% neem leaf from day 7 to day 35 gained less weight, used feed less efficiently, and developed raised lactate dehydrogenase, aspartate aminotransferase and alkaline phosphatase, raised uric acid and bilirubin, lower total protein and damage to liver and kidney; tissue recovery was incomplete two weeks after the leaf was withdrawn. Chicks.

167. 167
     Sutton NM, Bates N, Campbell A. Apparent adverse reactions to neem (margosa) oil in cats.Vet Rec · 2009 · 164(19):592-3doi:10.1136/vr.164.19.592 · PMID 19429937

     Veterinary poisons-service report of adverse reactions in cats to neem oil flea products; no abstract on PubMed. Included to show the oil's toxicity is not confined to human infants.

168. 168
     Gupta K, Jha M, Jadon RS, Sood R. Case of methaemoglobinaemia caused by tree oils and kerosene.BMJ Case Rep · 2017 · 2017:bcr-2017-220802doi:10.1136/bcr-2017-220802 · PMID 28942403

     A young man in India drank 100 mL of a household rat poison made of azadirachtin oil 40%, tea oil 15%, pine oil 25% and kerosene 20%; his methemoglobin fraction was 80%, he was treated with low-dose methylene blue and intravenous vitamin C, fell to 20% within an hour and was discharged after five days. One case; because the product contained kerosene and two other oils, neem cannot be isolated as the cause.

169. 169
     Subramaniam S, Ganesan A, Raju N, et al. Study of Respiratory Symptoms and Pulmonary Function among Neem Oil Industry Workers.Indian J Occup Environ Med · 2024 · 28(2):163-169doi:10.4103/ijoem.ijoem_250_23 · PMID 39114101

     50 men working in neem oil extraction in Tamil Nadu compared with 50 unexposed men: more cough, sneezing, wheeze and nasal irritation, and a fall in the ratio of forced expiratory volume in one second to forced vital capacity that tracked the duration of exposure. Cross-sectional, with exposure to dust and processing chemicals as well as to neem itself, so neem is not isolated.

170. 170
     Yerbanga RS, Rayaisse JB, Vantaux A, et al. Neemazal as a possible alternative control tool for malaria and African trypanosomiasis?Parasit Vectors · 2016 · 9:263doi:10.1186/s13071-016-1538-x · PMID 27146309

     NeemAzal in blood meals reduced feeding, longevity and host-odor response in Anopheles coluzzii and tsetse flies; in mice no death occurred after 14 days at oral doses of 3.8 to 12.7 g/kg, though behavior changed above 8 g/kg. Insects and mice.

171. 171
     Russo EB, Whiteley VL. Cannabinoid hyperemesis syndrome: genetic susceptibility to toxic exposure.Front Toxicol · 2024 · 6:1465728doi:10.3389/ftox.2024.1465728 · PMID 39507417

     Review that ties cannabinoid hyperemesis syndrome to escalating intake of high-potency cannabis in users carrying any of five reported mutations, and that explicitly addresses and refutes the alternative explanations, 'including exposure to pesticides, neem oil or azadirachtin'. Review; both authors are affiliated with CReDO Science, a cannabis research company.

172. 172
     Mishra MP, Padhy RN. In Vitro antibacterial efficacy of 21 Indian timber-yielding plants against multidrug-resistant bacteria causing urinary tract infection.Osong Public Health Res Perspect · 2013 · 4(6):347-57doi:10.1016/j.phrp.2013.10.007 · PMID 24524024

     Against nine multidrug-resistant urinary pathogens, neem leaf extract was in the group with 'moderate control capacity', behind ten other timber trees. Agar wells.

173. 173
     Gherbawy YA, AlOmari H, Al-Harthi HF, ElDewy E, Ioan P, Elhariry H. Sustainable control of aflatoxin B1, ochratoxin A, and fumonisin B1 in poultry feed using plant extracts and clay.Sci Rep · 2025 · 15(1):35966doi:10.1038/s41598-025-17693-9 · PMID 41093865

     2025: against Aspergillus and Fusarium in poultry feed, clove and cinnamon oils were strongest and 'neem extracts exhibited moderate efficacy'. In vitro and feed.

174. 174
     Fandohan P, Gbenou JD, Gnonlonfin B, Hell K, Marasas WF, Wingfield MJ. Effect of essential oils on the growth of Fusarium verticillioides and fumonisin contamination in corn.J Agric Food Chem · 2004 · 52(22):6824-9doi:10.1021/jf040043p · PMID 15506822

     Against the maize mold Fusarium verticillioides, lemongrass and basil oils inhibited growth completely; 'the oil of neem seeds showed no inhibitory effect but rather accelerated the growth'. In vitro and in stored grain; a plain negative for the antifungal claim.

175. 175
     Rodrigues MP, Astoreca AL, Oliveira ÁA, et al. In Vitro Activity of Neem (Azadirachta indica) Oil on Growth and Ochratoxin A Production by Aspergillus carbonarius Isolates.Toxins (Basel) · 2019 · 11(10):579doi:10.3390/toxins11100579 · PMID 31590398

     Neem oil (0.12% azadirachtin) inhibited 97.6% of fungal growth at 0.3% but only 40.2% at 0.5% and 64.7% at 1.0%, while blocking ochratoxin production at 0.1-0.3%. In vitro, and the dose-response goes the wrong way above 0.3%.

176. 176
     Ali J, Akram M, Zahoor M, Shah AB, Alotaibi A. Solvent-Dependent Bioactive Composition and Antifungal Potential of Neem (Azadirachta indica) Seed Extracts Against Mycotoxigenic Aspergillus Species.Chem Biodivers · 2026 · 23(2):e03312doi:10.1002/cbdv.202503312 · PMID 41631781

     2026: seed extracts in five solvents all inhibited Aspergillus at 500 mg/ml; ethanol extract cut fungal mass 45% at 31.25 mg/ml; methanol and ethanol extracts at 250 mg/ml suppressed aflatoxin over 90%; a six-month feed trial found seed powder at 30% blocked aflatoxin. In vitro and in feed, at very high concentrations.

177. 177
     Yu Y, Li Y, Gong Y. Nimbolide Reduces Inflammation and Oxidative Stress to Inhibit Disease Progression in Septic Acute Kidney Injury Mice.Arch Immunol Ther Exp (Warsz) · 2026 · 74(1)doi:10.2478/aite-2026-0023 · PMID 42632097

     2026: nimbolide reduced creatinine, urea, cytokines and tubular damage in mice with sepsis from cecal ligation, via NF-kappaB and STAT3. Mice.

178. 178
     Cai C, Zhang Z, Zhao X, et al. A novel METTL3 inhibitor nimbolide ameliorates osteoporosis via orchestrating osteoclastogenesis in an m6A-dependent manner.Phytomedicine · 2026 · 155:158048doi:10.1016/j.phymed.2026.158048 · PMID 41865687

     2026: nimbolide identified by docking, thermal shift and surface plasmon resonance as a direct inhibitor of the RNA methyltransferase METTL3, suppressing osteoclast formation and bone loss in ovariectomized mice. Mice and cells.