5-MeO-DMT
The largest reported effect size in the field, from the shortest session — and an indigenous tradition that appears to have been invented in 1984.
5-MeO-DMT is the field's outlier in almost every direction. Its acute experience is measured in minutes rather than hours; its principal molecular target is 5-HT1A rather than 5-HT2A, which is the target of licensed anxiolytics; it is the only psychedelic in development sourced — historically — from a vertebrate; and its Phase 2b results are, on their face, several times larger than psilocybin's Phase 3. It is also the compound whose cultural framing is most demonstrably false: the “ancestral toad medicine” narrative traces not to pre-Columbian practice but to a self-published pamphlet from 1984. This report separates the pharmacology from the mythology, and treats the extraordinary effect sizes with the scepticism their magnitude demands.
Its main target is 5-HT1A, not 5-HT2A.
Affinity is highest at 5-HT1A — the target of buspirone and a mediator of licensed antidepressant action. This makes it pharmacologically distinct from every other tryptamine in development.
The indigenous tradition appears to be manufactured.
There is no historical evidence for pre-1980s ceremonial use of Incilius alvarius. The practice traces to a 1984 pamphlet; the Comcáac ancestral framing was reconstructed afterwards.
It has an active metabolite on a polymorphic enzyme.
CYP2D6 O-demethylates it to bufotenine, which has ~10× higher 5-HT2A affinity than the parent. CYP2D6 is highly polymorphic — a real and under-discussed source of inter-patient variability.
GH001's Phase 2b is the largest effect in the field.
−15.5 placebo-adjusted MADRS at Day 8, 57.5% remission versus 0% on placebo, Cohen's d ≈ 2.0. Published in JAMA Psychiatry. Section V argues these numbers require unusual scrutiny.
Someone finally owns a molecule.
BPL-003 holds granted composition-of-matter patents in the US, UK and Europe on the benzoate salt — the one genuine exception to this field's universal “nobody owns the compound” problem.
It solves the session-length problem outright.
Vaporised effects resolve in roughly 20 minutes; most Phase 2b patients were discharged within an hour. No psychotherapy was used. This is the operational endpoint the whole field has been moving toward.
Ninety years, and a tradition invented halfway through
Origins:
plant first, toad second
5-MeO-DMT was made in a laboratory before anyone knew it occurred in nature. Toshio Hoshino and Kenya Shimodaira synthesised it in 1936, as part of structural work on serotonin-adjacent indolealkylamines — no psychoactive claim attached.1 That ordering matters, because it means the compound's pharmaceutical history does not depend on any ethnobotanical lineage.
The genuine traditional-use story is botanical, not amphibian. In 1959 5-MeO-DMT was identified as the major psychoactive constituent of yopo and cohoba — snuffs prepared from Anadenanthera seeds and used ceremonially across the Orinoco basin and the Caribbean for centuries.2 This is documented, pre-Columbian, and rarely mentioned in contemporary marketing, which prefers the toad.
The toad enters the record in 1965–67, when Vittorio Erspamer's group reported 5-MeO-DMT in the parotoid secretion of Incilius alvarius (formerly Bufo alvarius), the Sonoran Desert or Colorado River toad — the only vertebrate known to produce it, at concentrations reported between 20% and 30% of the secretion's dry weight.3 Erspamer was doing comparative biochemistry, not ethnography.
Human use in the West began, remarkably, as legal mail order. Through the early 1970s 5-MeO-DMT could be ordered in the United States from the Church of the Tree of Life and the Church of the Toad of Light. Microgram-accurate scales were not widely available, so it was commonly dispersed onto parsley or cannabis to make dosing tractable4 — a detail worth holding onto, because dose precision remains the central practical difficulty with a compound this potent and this fast.
Alexander Shulgin characterised it in TiHKAL, and it circulated in the same networks as the other tryptamines in this series. But 5-MeO-DMT never had a Sandoz — no Delysid, no Indocybin, no pharmaceutical sponsor until the 2020s. Its route to the clinic runs through the underground, not through a withdrawn medicine.
The tradition that was invented in 1984
This is the most important sociocultural fact about 5-MeO-DMT, and it runs directly against how the compound is usually presented. There is no historical evidence for indigenous ceremonial use of Incilius alvarius toads prior to the mid-1980s.5
The practice has a documented and recent origin. In 1984 Ken Nelson, writing under the pen name Albert Most, self-published a booklet titled Bufo Alvarius: The Psychedelic Toad of the Sonoran Desert, describing how to collect, dry and smoke the secretion. That pamphlet is the proximate cause of essentially all subsequent toad use.5,6
What followed is a well-documented case of retrospective traditionalisation. Scholarship — notably work reconstructing how a 1936 laboratory compound was transmuted into a “millenary practice” of the Comcáac (Seri) people of Sonora — traces how a for-profit wellness and retreat economy generated erroneous, coerced and exploitative narratives of ancient Indigenous use, with the effect of increasing exploitation of the toad, eroding actual biocultural knowledge, and licensing malpractice in 5-MeO-DMT administration.6,7
Psilocybin's programme carries a real and unpaid debt to María Sabina and the Mazatec. 5-MeO-DMT's carries an invented one — which is its own kind of harm, and a different one.
The contrast with psilocybin is instructive and should be drawn carefully. Psilocybin has documented pre-Columbian ceremonial use, a traceable line of transmission, and a genuine benefit-sharing obligation. 5-MeO-DMT's toad narrative has none of those things — but it has been used to make the same claims, attract the same reverence, and justify the same commercial arrangements. Anyone citing indigenous precedent for toad-derived 5-MeO-DMT is repeating a claim that the historical record does not support.
Incilius alvarius is now listed as imperiled or locally extinct across parts of its US range, under pressure from secretion harvesting for the retreat economy.4,7 Because the secretion is a defensive response released under stress or threat, some herpetologists hold that there is no humane collection method at all.5
Every clinical programme uses synthetic material, and this is the rare case where the pharmaceutical route is unambiguously the ethical one: a defined single-molecule product removes both the animal-welfare problem and the dose-uncertainty problem in one move. It is worth stating plainly, because the “natural versus synthetic” framing usually runs the other way in this field.
One empirical wrinkle deserves flagging rather than dismissing. Volunteers consuming toad secretion have reported subjective effects roughly 20–30% greater in magnitude than those given synthetic 5-MeO-DMT, prompting an “entourage molecule” hypothesis — that other constituents of the secretion modulate the experience.8 The comparison is naturalistic, uncontrolled, and confounded by expectancy and by dose uncertainty in the toad condition. It is not evidence that the secretion is superior; it is a hypothesis nobody has tested properly.
Mechanism: the 5-HT1A psychedelic
5-MeO-DMT is an agonist at both 5-HT1A and 5-HT2A, with highest affinity at 5-HT1A — an inversion of the classic psychedelic profile, and the single most consequential fact about its pharmacology.9,10 It also shows activity at 5-HT3A, 5-HT5, 5-HT6 and 5-HT7, with reported modulation of dopaminergic, glutamatergic and GABAergic signalling.10
Why the inversion matters: 5-HT1A is not a speculative target. It is the mechanism of buspirone, a contributor to the action of several licensed antidepressants, and a well-validated mediator of anxiolysis. A compound whose primary engagement is 5-HT1A with 5-HT2A secondary is, in principle, positioned differently from psilocybin or LSD — closer to an anxiolytic with psychedelic properties than a psychedelic with anxiolytic effects. Recent cryo-EM structures of 5-HT1A bound to bufotenine derivatives have begun to define the structural basis of that engagement, and explicitly frame 5-HT1A as the antidepressant-relevant target while attributing hallucinogenic effect to 5-HT2A.11
That framing carries an obvious implication the field is already pursuing: if therapeutic benefit runs through 5-HT1A and the overwhelming experience through 5-HT2A, then a 5-HT1A-selective analogue might separate them. The structural work is explicitly positioned as a framework for “next-generation nonhallucinogenic therapeutics.”11 Treat this as a well-motivated hypothesis with no human validation — the same status as the non-hallucinogenic 5-HT2A agonist programmes.
The metabolite problem
5-MeO-DMT is inactivated primarily by MAO-A deamination. But a second route matters clinically: CYP2D6 O-demethylates it to bufotenine (5-OH-DMT), which is itself psychoactive and carries roughly ten-fold higher 5-HT2A affinity than the parent compound.9,12 In mice, systemic exposure to bufotenine is about one-fourteenth that of 5-MeO-DMT, but its greater 5-HT2A potency means it may contribute disproportionately to the overall effect — and this metabolic conversion is the most economical explanation for why a nominally 5-HT1A-preferring compound produces a florid 5-HT2A-type experience.12
CYP2D6 is one of the most polymorphic drug-metabolising enzymes in humans, with poor, intermediate, extensive and ultra-rapid metaboliser phenotypes at substantially different population frequencies across ancestries — and it is inhibited by numerous common drugs, including several SSRIs. A compound whose active-metabolite exposure depends on CYP2D6 status should, in principle, show genotype-dependent variability in both effect and duration. We found no published clinical pharmacogenomic analysis of this in either sponsor programme. For a drug being dosed to fixed protocols in depressed patients frequently taking CYP2D6-inhibiting antidepressants, that is a real gap, not a theoretical one.
Kinetics and the peak
Route governs everything. Vaporised, effects begin within about 30 seconds and typically resolve within 20 minutes. Insufflated or intramuscular, onset is three to five minutes and effects subside within 45 to 60 minutes.13 Subjective effects include distortion of auditory and time perception, amplification of emotional state, and ego dissolution; individual dose escalation reliably induces a “peak” experience of the kind hypothesised to predict therapeutic response.9
Two consequences follow. Operationally, this is the shortest session in psychiatric development — the basis of the entire commercial thesis. Methodologically, it is also the least blindable intervention in the field: a 20-minute experience of complete ego dissolution is not something a participant can be uncertain about. Section V returns to what that does to a placebo-controlled trial.
Where 5-MeO-DMT sits among the tryptamines
The receptor inversion and the duration collapse are the two things that make this compound different. Rank order is directional, drawn from review-level sources rather than one harmonised binding panel.
The prodrug pattern, inverted. Psilocybin and luvesilocin are prodrugs that become the active 4-hydroxytryptamine. 5-MeO-DMT is the reverse case: it is active in its own right and its metabolite adds a second, more 5-HT2A-potent pharmacology on top — governed by a polymorphic enzyme. Where the 4-hydroxy programmes engineered away metabolic variability, 5-MeO-DMT carries it inherently.
The evidence base before the trials
5-MeO-DMT's pre-clinical-trial literature has an unusual shape, and it is worth being explicit about how weak it is. Unlike LSD and psilocybin — which have mid-century clinical literatures, however flawed — 5-MeO-DMT went almost directly from underground use to Phase 1, with surveys and naturalistic retreat studies standing in for controlled research.14
What that produced, in rough order of strength:
- Controlled human pharmacology (2019–22). Dose-escalation and characterisation work, synthesised in Reckweg et al.'s clinical-pharmacology review — the reference text for the compound, and the source of the dose-escalation/peak-experience framing the sponsors adopted.9
- Biomarker observations. Inhalation of synthetic 5-MeO-DMT in humans raised cortisol and lowered interleukin-6, with reduced stress and anxiety measures at seven days.10 Small and uncontrolled, but the IL-6 finding is mechanistically interesting given inflammation's role in depression.
- Preclinical plasticity and organoid work. Exposure in brain organoids dysregulated proteins associated with inflammation, long-term potentiation and cytoskeletal dynamics.10 Consistent with the psychoplastogen framing that covers psilocin, LSD and DMT — and no more than consistent.
- Retreat and survey studies. Naturalistic reports of large, rapid improvements in mood following ceremonial use. Uncontrolled, self-selected, expectancy-saturated, and — where toad secretion was used — of unknown dose. Hypothesis-generating at best.
The practical upshot: unlike every other compound in this series, 5-MeO-DMT's therapeutic case rests almost entirely on two sponsor-run Phase 2 programmes. There is no independent academic trial literature to triangulate against. That makes the effect sizes in Section V both more important and harder to sanity-check.
Two programmes, two comparators
Two companies are developing 5-MeO-DMT for treatment-resistant depression, by different routes and — crucially — against different comparators. That divergence turns the pair into something close to a natural experiment on the blinding question, and it is the most analytically useful feature of the whole landscape.
GH001 — inhaled, against inert placebo
GH Research's GH001 is an inhalable mebufotenin product. Its Phase 2b (GH001-TRD-201) was randomised, double-blind and placebo-controlled in 81 patients with TRD — 40 to GH001, 41 to placebo — with an individualised dose-escalation regimen and, notably, no psychotherapeutic component in either part of the trial.15
The results are the largest reported in this field. A placebo-adjusted MADRS reduction of −15.5 points at Day 8 (p < 0.0001), with a 57.5% remission rate versus 0% on placebo. Every secondary endpoint was met: CGI-S −2.5, HAM-A −10.0, Q-LES-Q-SF +21.4, all p < 0.0001. The MADRS anhedonia factor fell 9.9 points against 0.0 on placebo, Cohen's d = −2.00.15,16,17 In the open-label extension, 77.8% of completers were in remission at six months, with 63% having received only one to four treatments across that period.15 The Phase 2b was published in JAMA Psychiatry, and a subsequent analysis reported efficacy independent of prior treatment-failure count — Day 8 remission ranging 53.9–63.6% across patients with two to five-or-more lifetime failures, with no correlation between failure count and MADRS improvement (r = −0.13; p = 0.44).18 FDA lifted a US clinical hold on GH001 in January 2026.19
These numbers are four times psilocybin's Phase 3 difference, and they deserve scrutiny proportional to their size — which is a statement about evidentiary standards, not about the sponsor.
BPL-003 — intranasal, against a low active dose
AtaiBeckley's BPL-003 is a proprietary intranasal formulation of mebufotenin benzoate, delivered by a nasal spray device already used in an approved drug product.20 Its Phase 2b core study compared 8 mg and 12 mg single doses against a 0.3 mg low-dose active control — the same methodological device Compass used with 1 mg psilocybin and Reunion with its subperceptual comparator. Both active doses produced statistically significant and clinically meaningful reductions in depressive symptoms sustained to eight weeks, and 8 mg was selected for Phase 3.21
The open-label extension addressed retreatment directly: a 12 mg dose given eight weeks after the initial dose produced additional rapid, clinically meaningful antidepressant effects sustained for a further eight weeks — supporting continued and increased effect with repeat dosing.21 The four-part Phase 2a (NCT05660642) was published in the Journal of Psychopharmacology in March 2026, reporting mean Snaith-Hamilton Pleasure Scale scores falling from 8.4 at baseline to 1.5 at Day 85 — indicating an absence of anhedonia — with no serious adverse events and no withdrawals for adverse events.22 A fourth cohort testing a two-dose 8 mg + 8 mg induction regimen in patients on defined antidepressants has dosed its first patient, with data expected Q4 2026.22
Following a successful End-of-Phase-2 meeting announced in March 2026, FDA and AtaiBeckley aligned on a dual-trial Phase 3 programme with flexible dosing, aiming to demonstrate both efficacy and long-term safety. Initiation is on track for Q2 2026, with topline from both parallel pivotal studies guided for early 2029 and cash runway reaffirmed through that point. BPL-003 holds Breakthrough Therapy designation and is also in development for alcohol use disorder.20,23,24
Across all six reports in this series, the recurring structural fact is that nobody can patent the compound — psilocybin, LSD, DMT, MDMA, methylone and 4-OH-DiPT are all unpatentable as compositions, forcing every sponsor onto formulation, deuteration or prodrug claims. BPL-003 is the exception: it is covered by granted composition-of-matter patents in the US, UK and Europe on the benzoate salt, with further claims pending.20 If BPL-003 reaches approval it will have materially stronger exclusivity than any other asset discussed in this series — a point that tends to be lost in comparisons run purely on effect size.
A third, non-commercial programme deserves note: the Usona Institute — the non-profit also running a psilocybin Phase 3 — has a 5-MeO-DMT programme, paired with support for biocultural conservation of Incilius alvarius.25 Given the manufactured-tradition and conservation problems in Section II, a non-profit sponsor explicitly funding toad conservation is the closest thing to a considered response the field has produced.
GH001 and BPL-003 side by side
The comparator column is the one to read first. Effect sizes measured against inert placebo and against a low active dose are not interchangeable quantities.
The natural experiment. Two sponsors, one molecule, opposite comparator choices. GH001's enormous effect against inert placebo and BPL-003's more modest effect against 0.3 mg are, to a first approximation, the same drug measured two ways — and the difference between them is an estimate of how much a low active comparator absorbs. That is the most directly useful number in this entire series for anyone designing a psychedelic trial, and it is available almost nowhere else.
Safety, legal status, and what the brevity does not fix
The clinical safety record so far is genuinely clean: no serious adverse events in GH001's double-blind part or throughout its open-label extension, no evidence of treatment-emergent suicidal ideation, vital signs described as stable, no sedation or dissociative symptoms reported, and most patients discharged within an hour.15,16 BPL-003's Phase 2a likewise reported no serious adverse events and no withdrawals for adverse events.22 For a compound with this reputation, that is a meaningful result.
Four caveats keep it honest.
- The non-clinical record is worse than the clinical one. Adverse events in uncontrolled settings — including loss of consciousness, vomiting, and cases requiring intervention — are reported in the retreat and naturalistic literature, generally with unknown dose. That difference is the argument for a defined pharmaceutical product, not evidence that the molecule is benign.
- Brevity is not the same as gentleness. The 5–20 minute peak is described as among the most intense experiences in psychopharmacology. A short exposure with total incapacitation still requires continuous trained supervision, an airway-aware setting, and — because onset is under thirty seconds — no meaningful window to intervene before it begins.
- Drug interactions are structurally awkward. Inactivation runs through MAO-A, so MAOI co-administration is a serious concern. And because bioactivation to bufotenine runs through CYP2D6 — inhibited by several SSRIs — the interaction profile in a depressed population on existing antidepressants is more complicated than for the 4-hydroxytryptamines. BPL-003's Part 4 cohort dosing patients on defined antidepressants is the right study; it has not reported.22
- Cardiovascular and psychiatric exclusions do real work. As with every compound in this series, trials exclude psychotic and bipolar history. Nothing yet characterises risk in a less-selected population, and HPPD has not been systematically assessed for this compound at all.
Legal status. 5-MeO-DMT has been in US Schedule I since 2011, following a period in which it was uncontrolled at federal level and openly sold.26 It is controlled in most comparable jurisdictions. Unlike 4-OH-DiPT — which remains unscheduled and can therefore be handled without a Schedule I registration — any 5-MeO-DMT work requires the full licensing apparatus, and approval of either asset would require rescheduling.
On abuse liability: 5-MeO-DMT is not reliably self-administered, produces no withdrawal syndrome, and shows rapid tolerance — the classic-psychedelic profile rather than the entactogen one. Compulsive use is not a characteristic pattern. The Schedule I placement rests on the class-level assertion rather than on compound-specific dependence data.
The commercial landscape: psychedelics in Phase 2–3
Every psychedelic or psychedelic-adjacent compound in active late-stage psychiatric development. The two 5-MeO-DMT rows are highlighted — the shortest sessions in the table, and the only granted composition-of-matter position in it.
4-PO-DMT
same molecule
CYB003 / HLP003
lysergide D-tartrate
βk-MDMA
via luvesilocin
mebufotenin benzoate
mebufotenin
N,N-dimethyltryptamine
CYB004
midomafetamine
single enantiomer
How to read this table. The field has converged on shortening the session: intranasal, inhaled, buccal, subcutaneous and deuterated routes all exist to cut monitored clinic hours, the binding constraint on this entire class. 5-MeO-DMT sits at the end of that trajectory — there is nowhere shorter to go.
And one correction to the series' standing generalisation. The other five reports note that nobody in this field owns their molecule. BPL-003's granted composition-of-matter patents are the exception — a reminder that a novel salt form of an old compound can still be patentable subject matter, and that AtaiBeckley's position is therefore structurally stronger than a straight effect-size comparison suggests.
Established, probable, speculative
Grading is ours. A — multiple independent sources, consistent. B — solid primary evidence, limited replication. C — single source, sponsor-derived, or contested.
Established · A
- It is a 5-HT1A/5-HT2A agonist with highest affinity at 5-HT1A.9,10
- CYP2D6 O-demethylates it to bufotenine, ~10× more potent at 5-HT2A; MAO-A is the main inactivation route.9,12
- Synthesised 1936; identified in yopo/cohoba 1959; in I. alvarius 1965–67.1,2,3
- Vaporised effects begin in under a minute and largely resolve within ~20 minutes.13
- US Schedule I since 2011.26
- Toad-derived use traces to a 1984 pamphlet, not to documented pre-modern practice.5,6
Probable · B
- A single inhaled dose produces rapid, large reductions in depressive symptoms in TRD versus placebo.15,18
- Intranasal 8 mg and 12 mg beat a 0.3 mg active control to eight weeks.21
- Retreatment produces additional benefit and is well tolerated.21
- Acute clinical safety is good under supervision: no SAEs across both programmes' reported data.15,22
- Anhedonia specifically improves, and substantially.17,22
Speculative · C
- That a −15.5 MADRS difference survives a pivotal trial and a longer endpoint. The 0% placebo remission rate makes this the least safe extrapolation in the series.15
- Efficacy genuinely independent of prior treatment-failure count — one post-hoc analysis, one dataset.18
- That 5-HT1A rather than 5-HT2A carries the therapeutic effect, and that the two are separable.11
- CYP2D6 phenotype effects on response and duration. Unstudied clinically.
- The “entourage molecule” hypothesis for toad secretion.8
Research gaps, in order of consequence
GH001's −15.5 was measured against inert placebo at Day 8. Everything about this asset's value depends on what remains against a low active dose at a later endpoint — which is, fortunately, what BPL-003's Phase 3 will measure.
A drug whose active metabolite depends on the most polymorphic CYP enzyme, given to patients commonly taking CYP2D6-inhibiting SSRIs, with no published genotype-stratified analysis. This is the most obviously missing study in the programme.
A ketanserin-blockade study — the design that settled the question for LSD and psilocybin — would establish whether 5-MeO-DMT's benefit requires 5-HT2A or runs through 5-HT1A. It appears not to have been done, and it is cheap relative to what it would settle.
A 5-HT1A-preferring compound with a −10.0 HAM-A secondary result and no GAD programme is a conspicuous omission. On mechanism alone this is the psychedelic that should be in anxiety trials.
No academic group has run a controlled efficacy trial. Unlike psilocybin — where Hopkins, NYU and Imperial provide triangulation — 5-MeO-DMT's therapeutic case is entirely sponsor-generated. Usona's programme is the likeliest corrective.
The 77.8% six-month remission figure comes from OLE completers — a survivorship-selected group, with 22% early discontinuation upstream. Neither the true durability curve nor the optimal retreatment interval is established.
Sources
P1 peer-reviewed primary research or official instrument · P2 review, book or historical scholarship · P3 sponsor communication or trade press · P4 general reference, non-load-bearing.