AMT

AMT lasts a very long time, its onset varies by a factor of sixteen, and almost nothing one would want to know about it has ever been measured in a human being. It is not its potency that makes it dangerous, and nobody has measured that potency either. Nor does it act first of all like a classical psychedelic: alpha-methyltryptamine is 3-(2-aminopropyl)indole, the indole analogue of amphetamine, and its dominant documented mechanism is monoamine release, to which is added a reversible inhibition of monoamine oxidase, which makes it the only molecule in this set to carry its own MAOI inside itself.

Pharmacology

Monoamine release is the dominant layer. Blough et al., full text read, measure in rat synaptosomes a serotonin release EC50 of 21.7 ± 1.0 nM, and their prose fixes this attribution: compounds with an unsubstituted indole nucleus, the AMT family, release serotonin between 18.6 and 32.6 nM, and AMT also releases at the dopamine and noradrenaline transporters, but with far lower potency. The two other values on the same line, 78.6 ± 4.0 and 112 ± 6 nM, cannot be attributed: three readings of that table give three incompatible mappings, and only the offprint would settle it. What we hold is a hierarchy, not a pair of numbers.

5-HT2A agonism is real, and it is the route of the psychedelic component. Blough writes that all the releasing tryptamines in his panel, AMT included, are potent and efficacious 5-HT2A agonists in calcium mobilisation at the human receptor. In mice, pretreatment with M100907, a selective 5-HT2A antagonist, blocks the head-twitch response induced by AMT (Li 2025). The EC50 value of 23 nM that circulates comes from a table compiled on Wikipedia, a tertiary source.

MAO inhibition is established in principle, and not quantified. Arai et al. show that AMT inhibits MAO-A and MAO-B in rat brain, with a more marked selectivity for MAO-A, that the inhibition is reversible and the mode competitive. Wagmann et al. find it again on recombinant human MAO-A and in a human liver S9 fraction. Gey and Pletscher add a detail that is not trivial: in the rat, AMT is relatively more effective in the heart than in the brain, so the inhibition is not a cerebral phenomenon only.

But nobody can write that AMT is a potent MAOI. The only range actually read, in Wagmann's abstract, is 0.049 to 166 µM for a whole panel of fourteen compounds, that is three orders of magnitude, and no value specific to AMT appears in it. The figure of 380 nM one reads everywhere comes from Wikipedia, whose sole reference is that same article, whose full text exists only as a PDF that was not opened. It is plausible, it is not verified, and it counts because it is low enough for harder instructions to be drawn from it. Caution over interactions remains justified; certainty is not.

Metabolism forbids predicting anything at all. Malaca et al., full text read, identify seventeen human metabolites, note slow metabolism, and designate no cytochrome isoform, which makes any pharmacokinetic interaction of AMT unpredictable, with a CYP2D6 inhibitor such as diphenhydramine as much as with a CYP3A4 inhibitor. The question is not settled in a reassuring direction, it is not asked. The same authors observe that oxidative deamination, the major detoxication route of the N,N-dimethyltryptamines, was not detected, probably thanks to the alpha methyl, and they conclude from this that AMT has no need of a co-administered MAOI. Not detected, probably, over three hours of incubation and a single case: that is not a result.

Human pharmacokinetics: none. No half-life, no Tmax, no clearance, no imaging in any species whatever, and not a single neurotoxicity study that bears on AMT. The one ascribed to it concerns alpha-ethyltryptamine, another molecule, at eight times 30 mg/kg by the subcutaneous route in the rat: the transfer is illegitimate twice over, by the substituent and by the dosing regimen. It is to that same alpha-ethyltryptamine, Monase, that the story of the withdrawal for agranulocytosis belongs, documented for it and false for AMT.

What it does, and in how many people

Two bodies of data circulate, and they are wrongly set against each other as a small clinical series versus a large corpus of users. They are two small series, and the gap between them may be nothing but noise.

On the clinical side, the 1960s, no source read in the original: 20 mg by the oral route are said to produce an LSD-type reaction equivalent to about 50 µg of LSD, with effects appearing only three to four hours after ingestion, the majority of subjects reporting themselves nervous, irritable, restless, unable to relax. Visual effects are not prominent there and the experience is described as an unpleasant and long LSD reaction. All of this reaches us through a passage quoted from a 1975 book itself reproduced in a Wikipedia citation field, and the original articles have no abstract indexed anywhere: the human clinical file on AMT is, in practice, read by nobody in that chain.

On the user side, Wilcox retrospectively questioned fifteen people, a figure absent from his abstract and found again in Tittarelli's review. It governs the reading of everything else: 90% visual hallucinations, that is thirteen people. Two figures from the same passage are less copied and more useful than the visuals and the euphoria: 15% severely depressed mood the following day, and 35% anxiety, correlated with a crowded setting. Wilcox concludes that use could be dangerous in people predisposed to depression, and that is the most specific at-risk population statement in the whole corpus.

The field picture is mixed, stimulant and psychedelic: agitation and irritability, hyperthermia, anorexia, vomiting, inability to sleep, tachycardia, raised blood pressure, mydriasis, impaired reflexes and coordination, anxiety, nausea, moderately severe dysphoria. Of 55 cases of exposure to AMT alone reported to the British poison centre, 66% presented an acute mental health disorder, against 32% for mephedrone over the same period.

Duration, onset, dose

Nothing that follows is a dosage schedule. The gap between the Soviet antidepressant dose of Indopan, 5 to 10 mg, and the common dose in the user corpora, 15 to 30 mg by the oral route, is a factor of five to eight, and nobody knows where the boundary runs. The thresholds of 20 mg for euphoria, 30 mg for a strong hallucinogenic effect and 40 mg for a strong effect are attributed to Wilcox by Wikipedia: his abstract runs to five sentences and contains no milligram figure. The tiered dose bands come from a community wiki, with no primary source. The TiHKAL figures, which everyone cites at second hand, could not be read.

What is actually documented above that: 100 mg and 270 mg by the oral route produced severe poisonings from which both subjects came out without sequelae. So no imaginary cliff should be manufactured around 80 mg. The lethal threshold remains unknown, and two survivals at a high dose are not a guarantee.

Onset is the central trap. It runs from fifteen minutes, in the published case of a 100 mg ingestion, to three or four hours according to the 1960s clinical work and the 2015 review. A factor of sixteen, and the spread owes as much to the dose as to the route. Writing three to four hours as a fact produces a false certainty that kills: whoever has read that figure and feels something within fifteen minutes will conclude they took something else, whoever has read fifteen minutes and feels nothing after an hour will redose. Tittarelli puts it soberly: onset is rather delayed, hence a risk of further inappropriate doses that can lead to overdose.

Duration is the real structural danger. Twelve to twenty-four hours in general, up to two days in some subjects. That means premature redosing, exhaustion, dehydration, inability to sleep, and a drug interaction window whose length is unknown. Since MAO inhibition is reversible and competitive, it probably fades with the elimination of the molecule, so in hours and not in weeks; but since human elimination is unknown, so is the duration of the interaction risk. On tolerance, no experimental data: what circulates is a wiki applying the generic classical psychedelic pattern to AMT.

Harm reduction

Seizures are the hardest clinical signal we have. In the British poison centre series, 14% seizures out of 55 cases of AMT alone, against 2% out of 488 cases of mephedrone, odds ratio 9.35, 95% CI 3.26 to 24.18. This is passive pharmacovigilance, with its biases: mild cases do not call, the substance is not analytically confirmed, the user denominator is unknown. The signal remains strong, and it is probably the most actionable information on this page. A competing hypothesis deserves to be raised without being given as a result: the peak of calls in 2011 and 2012 coincides exactly with the European 5-IT wave, whose seizures are documented, and nobody has tested it. The point holds beyond this molecule. AMT is one of the ten substances in this corpus, out of fourteen, to carry a seizure signal, and it is among those that base it on counted human cases and not on an animal signal, an isolated case or an unquantified clinical mention. What belongs to it in its own right is the window: twelve to twenty-four hours during which adding another product from those ten amounts to stacking two lowerings of the seizure threshold. Tramadol and lithium are moreover flagged independently by several of these files.

What is sold as AMT may be 5-IT, and no reagent tells the difference. AMT and 5-IT, 5-(2-aminopropyl)indole, are positional isomers that are hard to tell apart in routine analysis. The 5-IT tally is clean and heavy: at most twelve deaths attributed to the poisoning out of nineteen deaths involving the product, plus fourteen non-fatal poisonings analytically confirmed, six of them severe with seizures and rhabdomyolysis. A field colorimetric test will not settle it, and the problem is wider: AMT, DPT and 5-MeO-DiPT are all indoles, therefore all Ehrlich-positive. The reagent separates none of these tryptamines from the others. Only an analytical test by an analysis service says anything.

The interactions rest on convergent mechanistic reasoning, and on zero published cases. AMT releases serotonin, inhibits its reuptake and inhibits MAO-A. Wagmann frames the consequence conditionally, and the conditional is in the source: MAO inhibition by the alpha-methylated tryptamines could be clinically relevant for serotonergic and adrenergic effects and for interactions with products acting as monoamine reuptake inhibitors. That is the right axis, and not a principle about releasers.

• MAOIs of every family, moclobemide, phenelzine, tranylcypromine, selegiline, linezolid, methylene blue, and within this set Syrian rue in all its forms, seeds, harmine, harmaline, ayahuasca. Adding an exogenous MAOI to a molecule that is already a serotonin releaser and a MAO-A inhibitor amounts to stacking the same risk twice. No published case exists, and the search was redone in order to be able to see the answer. This is an extrapolation, it must be given as such, and it is the one on which there is least to compromise. A point about ranking, which cuts both ways: in this set of fourteen, AMT and Syrian rue are the only two established monoamine oxidase inhibitors, and nutmeg, commonly credited with the same property, is not one. Its MAOI reputation rests on a 1963 article nobody has been able to read, with no published value and never replicated; the only modern measurement on the nutmeg side bears on another compound of the seed, macelignan, and on MAO-B, whereas the isoform that counts for serotonergic risk is MAO-A. That does not say that combining the two would be harmless, it says where the inhibition comes from: assumed for nutmeg, established by direct measurement for Syrian rue, whose potency is moreover quantified, and for AMT, whose potency is not.
• DXM is the most serious partner in the set. A SERT inhibitor as potent as fluoxetine at the brain concentrations reached in recreational use, it appears by name among the molecules involved in fatal serotonergic toxicities with MAOIs. It falls exactly on the axis Wagmann designates. No AMT plus DXM case has been published.
• SSRIs, SNRIs, tricyclics, tramadol, triptans, lithium, St John's wort: same axis. The support is not a case but the opinion, itself conditional, of a British drug information body.
• Opioids: do not write the word as a block, the distinction is actionable for someone on substitution treatment. Involved in serotonergic toxicities with MAOIs, deaths included: pethidine, tramadol, methadone, dextromethorphan, propoxyphene. Known not to be reuptake inhibitors and not to precipitate toxicity with an MAOI: morphine, codeine, buprenorphine, oxycodone. Caveat: this second half is not a general absolution, oxycodone being moreover among the opioids most reported in serotonin syndrome in the WHO database, through mechanisms independent of SERT.
• Other releasers and stimulants, MDMA, amphetamines, cocaine, cathinones: this is where the only clinical evidence lies. The only well documented European death is an AMT plus 5-MAPB co-intoxication, AMT 4690 ng/mL and 5-MAPB 101 ng/mL in peripheral blood, with benzodiazepines and ephedrine also present. The only published case of severe serotonin syndrome with AMT concerns tablets containing methoxphenidine and AMT; that report could not be read, its abstract does not exist in the database, and its clinical details are not verified. Serious events with AMT are combination events.
• The 2Cs of the set, 2C-I, 2C-E, 2C-T-7: an inference from two tables read, with no case at all. These three molecules have MAO-A and MAO-B as major elimination routes, and 2C-I has no documented back-up route. AMT inhibits MAO. The foreseeable risk is not first of all a serotonin syndrome, it is an overdose by blocked clearance on molecules active at a few tens of milligrams. Two inseparable caveats: since the MAOI potency of AMT is not quantified, this inference is weaker than the same one applied to Syrian rue; and for the 2C-T sub-family, the only published test that has looked at whether MAO blockade worsens toxicity is negative, on cellular cytotoxicity, which does not refute plasma accumulation but forbids giving the warning as a fact.
• 5-MeO-DiPT is the only other tryptamine in the set for which a MAO elimination route is established in humans, and extracts sold in Japan already contained, without declaring it, 5-MeO-DiPT together with harmine and harmaline. The team that compared AMT, 5-MeO-AMT and 5-MeO-DiPT in mice reports that AMT and 5-MeO-DiPT both show acute toxic effects.
• Vomiting while sedated is the most banal and the most lethal chain. Vomiting appears in the list of AMT adverse effects; the sedatives in this set are nitrous oxide, DXM and diphenhydramine, the last of these also circulating as an undeclared adulterant in powders sold under other names. Nothing is administered against this: what protects is somebody awake alongside, able to turn the person who vomits onto their side, and with this molecule that somebody has to hold out for twelve to twenty-four hours.
• Tyramine. A clinically significant MAO-A inhibition would in theory imply a risk of hypertensive crisis with foods rich in it. This has never been studied for AMT. Do not assert it, do not dismiss it.

Populations to exclude. People predisposed to depression, the only specifically documented statement and it comes from fifteen subjects. People on a serotonergic antidepressant, on an MAOI or on opioid substitution treatment. A history of epilepsy, by extrapolation from the seizure signal. Cardiovascular conditions. And hot, prolonged settings, festival, dancing, heat, where hyperthermia meets a duration of twelve to twenty-four hours.

What must send someone to the emergency department. The triad of serotonergic toxicity: altered consciousness, neuromuscular abnormalities (hyperreflexia, clonus, myoclonus, rigidity), autonomic hyperactivity (hyperthermia, profuse sweating, tachycardia, hypertension, mydriasis). Hyperthermia and rigidity tip it into a life-threatening emergency. Add to these a seizure, repeated vomiting with an inability to take in fluids, worsening agitation, dark urine with severe muscle pain. And a rise in blood pressure, with or without tachycardia: do not wait for the signs to come together, since in the 270 mg case blood pressure was 183/93 while heart rate was low at 52, temperature and respiratory rate normal.

The danger does not stop when the effects end. The published deaths do not have the same shape. In Miami, a 22-year-old man is found unresponsive about twelve hours after he was seen, iliac vein blood at 2.0 mg/L. In Southampton, a young man of 18 is admitted to hospital, placed in an induced coma and dies four days later of multiple organ failure. This second account comes not from a toxicological publication but from a pathologist's testimony at the coroner's inquest, reported in the press, with no blood concentration at all: the amount close to a gram attributed to him is an estimate of what would have been taken, not a measurement. The practical consequence holds all the same: a hospital admission for AMT justifies renal, hepatic and muscle monitoring over several days.

The death count, honestly. Three deaths with a quantified and published AMT concentration: Miami at 2.0 mg/L, a second medical examiner case in another jurisdiction at 1.5 mg/L in antemortem serum, and the French case at 4.7 µg/mL, published twice by two overlapping teams and amounting to one dead person only. Six British deaths counted for 2011 and 2012 with no published analytical data. One coroner's inquest death. Two men in 2015 according to a drug information body. Mentions without detail for the United States, Sweden, Norway and Japan. The overlap between these counts is not established: nobody knows how many distinct people this represents, and since the user denominator is unknown, none of these figures states a frequency. The WHO expert committee noted in 2014 that fatal poisonings involving AMT had been reported but that other products were present, and recommended not placing it under international control while keeping it under surveillance.

In the emergency department, nobody will be identified. No immunoassay screening test exists for tryptamine derivatives, and they do not appear in routine analyses: a negative screen rules nothing out. Conversely, diphenhydramine, present in several of the clinical pictures in this set, produces false positives, phencyclidine among them. Saying what was taken and at what time, and keeping the bag, materially changes management.

Sources

• B.E. Blough et al. (2014), interaction of psychoactive tryptamines with biogenic amine transporters and serotonergic receptors, Psychopharmacology (full text read; the column attribution of its table 1 is not reliable, three readings gave three mappings).
• R. Tittarelli et al. (2015), recreational use, analysis and toxicity of tryptamines, Current Neuropharmacology (full text read; source of the use doses, the duration, the 100 mg and 270 mg cases, and of Wilcox's sample size).
• S. Malaca et al. (2023), metabolite profiling of AMT in human hepatocytes and in post-mortem urine and blood, Metabolites (full text read).
• A. Kamour et al. (2014), clinical toxicity after reported use of AMT in the United Kingdom, Clinical Toxicology (abstract read; the most solid clinical source, all its figures verified).
• Y. Arai, Y. Toyoshima, H. Kinemuchi (1986), inhibition of monoamine oxidase by alpha-methylated monoamines, Jpn J Pharmacol and K.F. Gey, A. Pletscher (1962), alpha-alkylated tryptamine derivatives and 5-HT metabolism in vivo, Br J Pharmacol Chemother (abstracts read; reversibility, competitive mode, short duration of action, heart more than brain).
• L. Wagmann et al. (2017), in vitro monoamine oxidase inhibition potential of AMT analogues, Toxicology Letters (abstract read only, full text a PDF that was not opened: it is the only route to a value specific to AMT, and therefore to the unverified 380 nM).
• F. Nagai, R. Nonaka, K. Kamimura (2007), psychoactive drugs and monoaminergic neurotransmission, Eur J Pharmacol and R. Nonaka et al. (2007), in vitro screening by GTPgammaS binding, Biol Pharm Bull (abstracts read; the second does not measure 5-HT2A).
• K. Li et al. (2025), three tryptamines and acute toxicity, locomotor activity and hallucinogenic behaviour in mice, Behavioural Pharmacology (abstract read; blockade of the head twitch by M100907).
• D.M. Boland et al. (2005), deaths from acute AMT intoxication, J Anal Toxicol (abstract read; two cases, 2.0 mg/L and 1.5 mg/L. The one-gram bottle found at the scene was empty and the gastric contents carried 9.6 mg in total: the ingested dose is unknown).
• M. Epain et al. (2024), fatal AMT and 5-MAPB co-intoxication, Int J Legal Med (abstract read; the same death as the one described by Malaca 2023, duplicate confirmed).
• J. Wilcox (2012), psychoactive properties of AMT from user reports, J Psychoactive Drugs (abstract read, full text not read. n = 15, sample size absent from the abstract, found again via Tittarelli. His abstract contains no dosage).
• H. Long, L.S. Nelson, R.S. Hoffman (2003), AMT revisited by easy internet access, Vet Hum Toxicol (abstract read; 100 mg ingestion, symptom onset at fifteen minutes).
• P.K. Gillman (2005), MAOIs, opioid analgesics and serotonin toxicity, Br J Anaesth (abstract read; opioid-by-opioid distinction. This abstract does not mention MDMA and does not speak of releasers: any citation of Gillman on that subject would be false).
• S.P. Elliott et al. (2013), AMT and 5-IT, an analytical challenge, Drug Testing and Analysis, R. Kronstrand et al. (2013), J Anal Toxicol, L.N. Seetohul, D.J. Pounder (2013), J Anal Toxicol and M. Bäckberg et al. (2014), the STRIDA project, Clinical Toxicology (abstracts read; the 5-IT record).
• X.M. Huang, M.P. Johnson, D.E. Nichols (1991), reduction of brain serotonergic markers by alpha-ethyltryptamine, Eur J Pharmacol (bears on alpha-ethyltryptamine, not on AMT, at eight doses of 30 mg/kg).
• WHO, Expert Committee on Drug Dependence, 36th meeting (2014), item 4.20 (HTML page read; the critical review report and the peer reviews are PDFs that were not opened).
• DrugWise, AMT page, updated January 2022 (British drug information body, not a publication; source of the 2015 deaths and of the statement on SSRIs and SNRIs).
• B. Chrétien et al. (2018), severe serotonin syndrome in an autistic user after consumption of tablets containing methoxphenidine and AMT, J Clin Psychopharmacol 38(1):94-96, PMID 29210865. NOT READ: the abstract does not exist in the database. A French article in Toxicologie Analytique et Clinique describing a very similar presentation returns a 403 and could be the same patient published twice.
• H.B. Murphree et al. (1961), Clin Pharmacol Ther 2(6):722-726; H.E. Himwich (1961); M.D. Mashkovskii and T.K. Trubitsyna (1963), the primary source on Indopan, in Russian. NONE OF THEM HAS AN INDEXED ABSTRACT, none was read. The whole human clinical corpus on AMT reaches us through a passage quoted from a 1975 book itself unread, and the head-twitch ED50 of 7.9 mg/kg attributed to Corne and Pickering 1967 crosses three tiers of citation without any original ever being shown.
• A. Shulgin and A. Shulgin (1997), TiHKAL, entry 48, and Erowid, AMT Vault. NOT READ (HTTP 403). The TiHKAL dose and duration figures do not appear here for want of being able to read them. Erowid is cited and linked, never copied out: its licence forbids republication of its corpus, and an experience report is referenced, it is not reproduced.
• English Wikipedia, α-Methyltryptamine and PsychonautWiki, αMT. Tertiary sources, used as an index of primary references. Two of their claims are explicitly set aside here: the MAO-A IC50 of 380 nM, and the thresholds of 20, 30 and 40 mg attributed to Wilcox.