Research journey
MDMA: 141 trials, 344 papers, one canvas
Patented in 1912, criminalised in 1985, and rejected by the FDA in 2024 after the most complete filing psychedelic medicine has ever assembled. 139 trials, 336 papers, two positive Phase 3 studies and no approval.
Milestones are curated from the Blossom research database. Registry completion dates are shown as planned; probability-weighted forecasts are part of Blossom Enterprise.
Hover any marker to preview its callout on the canvas, then click to pin it. Pinned is what stays open, fills this panel with the full context and links, and is what an export captures. Landmarks stay labelled on the canvas so the exported image reads on its own.
Download the artifact
Exports rasterise the live SVG, so whatever you have toggled is what you get.
Origins
1912 to 1984Merck's patent, Shulgin's revival, the therapist underground.1912
Merck
Merck synthesises MDMA
The molecule is made at Merck in Darmstadt as an intermediate, patented, and then left alone for six decades.
MDMA is first synthesised at Merck in 1912 and appears in a 1912 patent filing as a chemical intermediate rather than a candidate drug. Nobody was looking for a psychiatric medicine. They were looking for a haemostatic (a compound that stops bleeding), and this was a step on the way.
What follows is the longest dormancy of any compound on Blossom. Sixty-four years pass before anyone gives MDMA to people on purpose, and a full century passes before it enters a Phase 3 trial. The persistent story that Merck developed it as an appetite suppressant is not supported by the company archive.
That gap matters for how you read the rest of this canvas. Everything below is compressed into the last fifty years, and most of the clinical evidence into the last twenty.
1976
Shulgin resynthesises MDMA
Alexander Shulgin makes the compound again, tries it, and passes it to a small circle of psychotherapists.
Alexander Shulgin resynthesises MDMA in 1976 after a student mentions its effects, then introduces it to the psychologist Leo Zeff, who introduces it to several thousand therapists over the following decade. There is no company, no protocol and no registry. There is a chemist, a network and a lot of unrecorded clinical practice.
This is the origin of a problem the FDA will name almost fifty years later. MDMA arrived in psychiatry as a therapy adjunct with an established underground reputation, which means the people who eventually enrolled in its trials were unusually likely to have taken it before and unusually likely to know which arm they were in.
David Nichols later gave the class its name, arguing in 1986 that MDMA and its relatives act differently enough from the classic hallucinogens to warrant their own category: entactogens. The taxonomy still matters, because it is why MDMA sits on a different regulatory and mechanistic track from psilocybin.
1980-01
Greer & Tolbert
Legal therapy, before prohibition
George Greer and Requa Tolbert run an open clinical series with 29 people while MDMA is still uncontrolled.
Between 1980 and 1983, before any scheduling decision, Greer and Tolbert administer MDMA in a therapeutic setting and write up the subjective reports. It is the earliest trial record in Blossom for this compound, and it exists because the legal question had simply not been asked yet.
The design would not pass an ethics committee today. There is no control group, no blinding and no standardised outcome measure, and the participants were selected by the therapists. Read it as documentation of practice rather than as evidence of efficacy.
Its real value is chronological. It shows that supervised clinical use of MDMA predates the ban, which is the fact the whole 1985 scheduling argument turned on.
Prohibition and persistence
1985 to 2016Schedule I, MAPS, and the slow build of evidence.1985-07
ended earlyUS emergency scheduling
The DEA places MDMA in Schedule I on an emergency basis, over the objection of its own administrative law judge.
The DEA uses emergency scheduling powers in July 1985 to place MDMA in Schedule I. In the hearings that follow, the agency’s own administrative law judge recommends Schedule III, which would have preserved prescribing. The DEA rejects the recommendation, and the placement is confirmed permanently in 1988.
This is the single most consequential decision on the canvas and it is worth being precise about what it did. It did not establish that MDMA is dangerous. It ended a legal therapeutic practice, imposed a licensing and storage burden that most university departments would not carry, and pushed the compound into a research category where every subsequent study needed federal permission.
The cost is visible in the shape of this timeline. Between 1985 and 2000 there is one strand of activity, healthy-volunteer pharmacology in Europe, and no patient research at all. Fifteen years of a fifty-year story disappear here.
1986
MAPS
MAPS is founded to fight the ban
Rick Doblin sets up a non-profit whose explicit purpose is to make MDMA a prescription medicine.
The Multidisciplinary Association for Psychedelic Studies is founded in 1986, the year after the ban, with a single strategic objective: get MDMA approved. Everything on the MDMA-AT lane below is downstream of that decision, funded almost entirely by philanthropic donations rather than venture capital.
The model is genuinely unusual and it shaped the science. A non-profit sponsor has no investors demanding a quarterly readout, which is how a programme survived thirty-eight years between founding and filing. It also means the organisation that designed the trials was the organisation that had spent decades advocating for the result, and the FDA advisory committee said so out loud in 2024.
Hold both of those at once. The patient advocacy is why this compound reached a regulator at all, and the advocacy is also part of why the filing failed.
1998-10
University of Zurich
First controlled healthy-volunteer study
Swiss researchers characterise the psychological and cardiovascular effects of MDMA in people who had never taken it.
Franz Vollenweider’s group publishes the first modern controlled administration study in MDMA-naive healthy volunteers, covering both the psychological profile and the cardiovascular response. It has been cited around 450 times and it is the foundation of everything on the human pharmacology lane.
The cardiovascular half is the part that ages best. MDMA raises heart rate and blood pressure reliably, which is unremarkable in a screened twenty-five-year-old volunteer and is exactly the question a regulator asks about a fifty-year-old with treatment-resistant post-traumatic stress disorder. That question was still open in 2024.
2000-01
MAPS
ended earlyThe first patient trial is shut down
A MAPS-funded low-dose pilot in Madrid for women with chronic PTSD is terminated before it can finish.
José Carlos Bouso runs the first controlled MDMA trial in patients anywhere, six women with chronic post-traumatic stress disorder after sexual assault, at low doses. It is terminated in 2002 under political pressure on the hospital, and the results are only published in 2008.
The trial ended for reasons that had nothing to do with what it found. That is worth stating plainly, because the canvas contains several stopped studies and the registry almost never distinguishes between a safety signal, a funding failure and an administrative decision taken by someone who never read the protocol.
It also set the pattern for the next two decades. MAPS-funded work is repeatedly the first of its kind, small, and conducted in whatever jurisdiction will permit it.
2001-06
University of Basel
Basel opens its interaction programme
A pindolol interaction study begins two decades of systematic human pharmacology at the University of Basel.
The Basel group starts working through MDMA’s pharmacology one interaction at a time: pindolol, reboxetine, duloxetine, clonidine, carvedilol, doxazosin, bupropion, methylphenidate. Twenty-five years later the same lane is still producing studies, now on the separated enantiomers (the two mirror-image forms of the molecule) and on analgesia.
This is unglamorous work and it is why the compound has a usable safety file at all. When a regulator asks what happens if a patient on an antidepressant is dosed with MDMA, the answer comes from this lane rather than from any Phase 3.
It is also independent of the approval effort. Basel has no product, so its results arrive in the literature whichever way the commercial story goes.
2004-03
MAPS
The first US patient trial
Michael and Annie Mithoefer open the randomised pilot in Charleston that produces the field’s founding efficacy result.
Twenty patients with chronic treatment-resistant post-traumatic stress disorder, randomised between MDMA-assisted therapy and an inactive placebo, run over six years in a single South Carolina practice. The 2010 publication reports that most people in the active arm no longer met diagnostic criteria, and it has been cited more than 650 times.
This is the number that built the field. Every later trial, every designation and the entire filing rest on the claim that a handful of supervised sessions can do what months of standard treatment often cannot.
Two caveats travelled with it from the start and were never resolved. Twenty patients is a pilot, and comparing an intense psychoactive experience against an inert placebo means almost everyone knew which arm they were in. A 2012 follow-up reported that the improvements held for years and found no evidence of dependence, which strengthened the durability claim considerably. The FDA still considered the durability evidence insufficient fourteen years later.
The sprint
2017 to 2023Breakthrough status, two positive Phase 3 trials, Australia first.2017-08
MAPS
FDA Breakthrough Therapy Designation
The regulator agrees MDMA-assisted therapy may substantially improve on existing PTSD treatment and commits to closer guidance.
Breakthrough Therapy Designation is the FDA signalling that preliminary evidence suggests a substantial improvement over available therapy, in exchange for more intensive engagement through development. MDMA gets one for post-traumatic stress disorder in August 2017, a year before psilocybin gets its first.
It is easy to read a designation as a promise. It is not. It is a commitment to talk more often, granted on early data, and the FDA can and does still refuse the eventual filing. Seven years later it did exactly that.
2018-06
Lancet Psychiatry
Dose-response trial in first responders
A randomised dose-response Phase 2 in veterans, firefighters and police officers publishes in Lancet Psychiatry.
Twenty-six participants, three MDMA doses, blinded and randomised. The higher doses outperformed the lowest, which is the kind of internal dose-response gradient that makes a drug effect more believable than a simple drug-versus-placebo comparison does.
The population is the strategic point. Veterans and first responders are the constituency that gave MDMA its political coalition in Washington, and the reason a compound criminalised in 1985 could be discussed sympathetically in Congress by 2020.
2018-11
MAPS Public Benefit Corp
MAPP1, the first psychedelic Phase 3
The first pivotal trial of any psychedelic-class medicine opens across the US, Canada and Israel.
MAPP1 is the first Phase 3 trial of a psychedelic-class compound anywhere. It opened thirty-two years after MAPS was founded and thirty-three years after the ban, funded by donations rather than by an investor round.
Reaching a pivotal trial at all was the achievement. Whether the design could survive regulatory scrutiny was a separate question, and the answer arrived six years later.
2021-05
Nature Medicine
MAPP1 publishes in Nature Medicine
The first pivotal psychedelic trial reports a large effect in severe PTSD. It is the most-cited MDMA trial in Blossom.
Ninety participants with severe post-traumatic stress disorder, randomised, double-blind, placebo-controlled. The MDMA arm improved substantially more than placebo on the standard clinician-administered severity scale, with no serious adverse events attributed to the drug. The paper has been cited more than 940 times.
On the day it published, this looked like the end of the argument. A criminalised compound had produced a positive pivotal result on the most difficult psychiatric indication in the book, and approval seemed a formality.
The design carried one weakness that the paper itself acknowledged and that no subsequent trial fixed. In the active arm, participants and therapists could almost always tell what had been administered, which means the measured effect includes an unknown quantity of expectation. A 2026 systematic review of blinding integrity across psychedelic trials put numbers on how large that problem is across the whole field.
2023-05
atai Life Sciences
EMP-01 first-in-human
A single-ascending-dose Phase 1 opens for R-MDMA, the single-enantiomer version of the molecule.
EMP-01 is R-MDMA: one of the two mirror-image forms of the molecule rather than the equal mixture used in every trial above. The rationale is that the R form may carry the prosocial effects with less stimulation and less cardiovascular load, and preclinical work in mice supports that.
Single-enantiomer development is also a patent strategy. Racemic MDMA is a century-old compound that nobody can own, whereas a purified enantiomer with its own clinical data is a defensible asset. Read the two commercial lanes on this canvas with that in mind.
2023-07
Therapeutic Goods Administration
Australia reschedules MDMA
Authorised psychiatrists can prescribe MDMA for PTSD, more than a year before any regulator rules on a product.
From 1 July 2023 the Australian Therapeutic Goods Administration permits specifically authorised psychiatrists to prescribe MDMA for post-traumatic stress disorder, alongside psilocybin for treatment-resistant depression. Australia becomes the first country in the world to allow it.
This is a different theory of access from the one the whole rest of this canvas assumes. There is no approved product, no marketing authorisation and no agreed label. There is a named prescriber who has satisfied an ethics committee and the regulator, treating one patient at a time.
It turned out to matter more than anyone expected, because thirteen months later the FDA route closed. Since August 2024, Australia has been the only jurisdiction where MDMA-assisted therapy is lawfully available as a treatment rather than as a trial, and the data now accumulating there is real-world rather than randomised.
2023-09
Nature Medicine
MAPP2 confirms the result
The second pivotal trial reports in Nature Medicine, in a broader moderate-to-severe population.
MAPP2 enrolled 104 participants with moderate to severe post-traumatic stress disorder and reproduced the MAPP1 finding in a wider population. Two positive pivotal trials is the conventional threshold for a US filing, and this is the point at which the company had a package.
A separate open-label extension study, for participants who wanted further sessions, completed in the same month. Both records sit in Blossom.
Replication of an effect does not replicate away a design flaw. Both studies compared MDMA against an inactive placebo, so both inherit the same unblinding problem, and running a study twice does not resolve a criticism that applies to how it was run.
2023-10
Portland VA Research Foundation
The VA runs its own group trial
A Portland VA study of group MDMA therapy for veterans completes, a year after the federal rejection.
Group-MVP tested MDMA-assisted therapy delivered to veterans in groups rather than one to one, and completed in September 2025. Group delivery attacks the cost problem directly: the therapist time around each session is the single largest expense in this treatment model, and one clinician cannot supervise one patient for eight hours at scale.
The sponsor is a Veterans Affairs research foundation, not a drug company. Public research kept going after the rejection because the population it serves did not change, and the academic and veteran lane on this canvas is now busier than the commercial one.
2023-12
Lykos Therapeutics
NDA 215455 is filed
The first ever new drug application for a psychedelic-class medicine is submitted and granted Priority Review.
Lykos Therapeutics submits NDA 215455 for midomafetamine capsules, used with psychological intervention, for post-traumatic stress disorder. The FDA accepts the filing and grants Priority Review, shortening the decision clock.
Thirty-eight years separate the ban from the filing. No psychedelic-class compound had ever got this far, and the sequencing on this canvas is worth noting: MDMA reached a regulator years ahead of psilocybin, which is why its rejection reset expectations across the entire field rather than just for one company.
The reckoning
2024 onwardsAn FDA rejection, and what the field does with it.2024-06
U.S. Food and Drug Administration
ended earlyAdvisory committee votes 9-2 against
The FDA’s Psychopharmacologic Drugs Advisory Committee recommends against approval, citing unblinding, data integrity and cardiovascular risk.
The advisory committee votes 9 to 2 against recommending approval. The reasons given were functional unblinding, data integrity concerns, cardiovascular safety, and the high proportion of participants who had used MDMA before enrolling.
Every one of those objections is visible earlier on this canvas. The unblinding problem is inherited from the 2004 pilot design and carried unchanged through both Phase 3 trials. The prior-use problem traces back to 1976, when the compound entered psychiatry through a therapist network rather than through a laboratory.
Advisory committees are advisory. The FDA is not bound by the vote, and it approves against committee recommendations from time to time. It did not here.
2024-08
U.S. Food and Drug Administration
ended earlyFDA issues a Complete Response Letter
The application is refused. The regulator asks for a new Phase 3 trial, and the company cuts most of its staff.
On 9 August 2024 the FDA issues a Complete Response Letter for NDA 215455. A complete response is a refusal: the application cannot be approved in its current form. Three deficiencies were cited. Adverse events with a positive valence, the ones that bear on abuse potential, were not systematically captured. The durability evidence was insufficient. And the proportion of participants with prior MDMA experience was too high to interpret the results cleanly.
The remedy the FDA asked for was a new Phase 3 trial. That is not a paperwork fix. It is several years and a large amount of money, demanded of a company whose entire value was the application that had just been refused.
The fallout was immediate. Lykos cut around three quarters of its staff, its chief executive departed, and Rick Doblin left the board of the company his non-profit had created. In the same month three MDMA papers were retracted from Psychopharmacology over unreported alleged misconduct at a Phase 2 site, which sharpened the data-integrity criticism considerably.
This is the most important marker on the canvas and the honest reading of it is uncomfortable in both directions. Two positive pivotal trials were not enough, which tells you something real about how weak an unblinded psychiatric trial is as evidence. It also left an indication with few effective options no closer to a new treatment, and a field that had told itself approval was inevitable had to stop saying so.
2024-11
JAMA Network Open
The hyponatraemia mechanism, pinned down
A pooled secondary analysis of four randomised trials shows how MDMA causes dangerously low blood sodium, and how to prevent it.
MDMA raises oxytocin and vasopressin, which makes the body retain water. Combine that with the fluid intake that recreational users are advised toward and you get hyponatraemia (blood sodium falling to dangerous levels), which is the mechanism behind most MDMA deaths outside overdose.
Pooling four randomised trials isolates the effect and shows fluid restriction during a session substantially reduces it. This is the sort of finding that only comes out of controlled research, and it makes supervised use safer whether or not any product is ever approved.
Published three months after the rejection. Safety science does not stop when a filing fails, and this lane is one of the few places on the canvas where the record has kept improving since 2024.
2025-05
AtaiBeckley
EMP-01 Phase 2a doses its first patient
Seventy-one adults with social anxiety disorder in the UK, two doses four weeks apart, and no psychotherapy component.
The design is a deliberate break from everything on the MDMA-AT lane. There is no manualised psychotherapy wrapped around the dosing, which means the trial tests the drug rather than the drug plus a therapy protocol.
That distinction is the strategic lesson the sector took from the rejection. A drug application that includes a psychological intervention has to persuade a regulator about the therapy as well as the molecule, and the FDA does not regulate psychotherapy. Removing it makes the filing simpler, and makes any positive result a narrower claim.
Social anxiety disorder is also an under-served indication with a large population, and the earlier autistic-adults work on this canvas suggested MDMA might help there.
2025-07
ended earlyTwelve trials that never delivered
Of 139 registered MDMA trials, five were terminated and seven withdrawn before enrolling anyone.
Withdrawn means a study was registered and then closed without enrolling a single participant. Terminated means it started and stopped early. Twelve of the 139 MDMA trials in Blossom carry one of those statuses, and the list runs across exactly the institutions you would expect to finish: Yale withdrew an imaging study in patients with post-traumatic stress disorder, a multi-site eating-disorder study was withdrawn, a veterans group-therapy feasibility study was withdrawn, and in July 2025 UCSF withdrew a study in adolescents.
The adolescent withdrawal is the one worth pausing on. Post-traumatic stress disorder in young people is badly served, the study was registered after the rejection, and it closed without a participant.
Registered intent and delivered research are different quantities. A pipeline chart counts the first and implies the second.
2025-08
Resilient Pharmaceuticals
Lykos becomes Resilient Pharmaceuticals
A year after the rejection, the company rebrands under new leadership and commits to a fresh Phase 3.
In August 2025 Lykos Therapeutics renames itself Resilient Pharmaceuticals, with a new chief executive and chief medical officer, and states that it will continue developing MDMA-assisted therapy for post-traumatic stress disorder.
This is the fourth name on one lane. MAPS raised the money, MAPS Public Benefit Corp ran the trials, Lykos filed the application, and Resilient inherits the deficiencies. The custody chain is the story: an asset that has changed hands three times without ever changing owner in the way an acquisition would.
The programme status in Blossom is paused. That is the accurate word. There is no registered trial for the study the FDA asked for, and until one appears the correct thing to say is that the leading MDMA programme in the world is not currently running a pivotal study.
2025-12
Definium Therapeutics
MM402 opens in autism
A second R-MDMA programme starts a Phase 2a in autistic adults, with topline guided for late 2026.
Definium Therapeutics completed a Phase 1 single-ascending-dose study of MM402, also R-MDMA, in October 2024 and opened a Phase 2a in autism spectrum disorder in December 2025. Topline is guided for the fourth quarter of 2026.
Autism is a contentious indication and the framing deserves care. The target is social anxiety and social difficulty in autistic adults, not autism itself, and a substantial part of the autistic community objects to research framed as treating the condition. The earlier randomised pilot in autistic adults with social anxiety, published in 2018, is the evidence this builds on.
Two independent companies pursuing the same enantiomer in adjacent social-anxiety indications is the clearest signal on the canvas of where capital went after August 2024. It went to the parts of MDMA that can be patented and tested without a therapy protocol attached.
2026-02
AtaiBeckley
EMP-01 Phase 2a reads out
The safety endpoint was met. The efficacy signal is modest and exploratory: an 11.85-point scale reduction, one-tailed.
The primary endpoint was safety and it was met, with no serious adverse events and no suicidal ideation reported. The efficacy result was exploratory: an 11.85-point reduction on the Liebowitz social anxiety scale, a Hedges’ g effect size of 0.45, and a p value of 0.036 calculated one-tailed.
Those qualifiers matter. Exploratory means the study was not designed to prove this, and a one-tailed test is a weaker standard than the two-tailed convention. A moderate effect from a study that was not powered to detect one is a reason to run a larger trial, not a result.
Stated plainly and early, which is the right way to read a topline. The company is treating it as support for a Phase 2b, and that is what it supports.
2026-03
Living systematic review
A living meta-analysis with open data
The MDMA-for-PTSD evidence base is pooled into a continuously updated review with the underlying data published.
A living systematic review updates as new trials appear, rather than freezing at publication. Pairing it with an open data resource means anyone can re-run the analysis with different assumptions rather than arguing about the abstract.
This is a direct response to the data-integrity criticism. The strongest available answer to a regulator who does not trust how results were reported is to publish the inputs, and until 2024 this field mostly did not.
2026-04
JAMA Psychiatry
Blinding integrity, measured
A systematic review and a head-to-head trial confirm that participants in psychedelic trials almost always know which arm they are in.
Two 2026 papers put numbers on the objection that sank the filing. A systematic review of blinding integrity across psychedelic randomised trials in April, and in May a randomised comparison of psilocybin, MDMA and methylphenidate in healthy volunteers designed specifically to test whether an active comparator can preserve the blind.
The methylphenidate arm is the interesting part. If a stimulant that produces noticeable but different effects can stop participants correctly identifying MDMA, then a future Phase 3 has a design that answers the FDA. If it cannot, then the honest conclusion is that this class of compound may not be testable to the standard the regulator applies to everything else.
That is the question the entire right-hand side of this canvas depends on, and it is a methodological question rather than a pharmacological one. Nineteen months after the rejection, it is being studied properly for the first time.
2026-06
SAMATI
Social anxiety results publish
An academic randomised trial of MDMA-assisted therapy in social anxiety disorder reports, open-label and wait-list controlled.
The SAMATI study publishes in June 2026: randomised, but open-label and controlled against a wait list rather than against a placebo. A wait-list control tells you how much better treated participants did than untreated ones, which includes everything about being in a trial and receiving attention.
It is the weakest of the common control designs and it is being used here because it is honest about what it can show. After 2024, small academic groups are less inclined to claim more than their design supports.
It also lands in the same indication as both commercial R-MDMA programmes, three months after one of them reported. Social anxiety is quietly becoming the second front for this compound.
2026-10
U.S. Army Medical Research and Development Command
plannedThe US Army runs its own trial
A Phase 2 in serving military personnel, sponsored by US Army Medical Research and Development Command, is due to open in October 2026.
Two years after the FDA refused the application, the US Department of Defense is opening its own MDMA trial in serving service members. The registry lists a start in October 2026 and completion in October 2028.
Government sponsorship changes the incentives in a way worth naming. The Army has no product to sell and no share price, which removes the commercial motive that the advisory committee treated as a source of bias. It also has a large population with post-traumatic stress disorder and few good options, which is why it is here.
Whether it can solve the blinding problem the FDA raised is unknown. Nothing in the registry record suggests a novel control design.
2027-06
Resilient Pharmaceuticals
plannedThe Phase 3 the FDA asked for
Resubmission requires a fresh pivotal trial. Almost two years after the rejection, none has been registered.
This marker has no registry entry behind it, which is precisely the point. The FDA asked for a new Phase 3 in August 2024. As of July 2026 no such study appears in Blossom, and the programme is recorded as paused.
The date shown is an illustration of the earliest plausible start rather than a forecast, and it should be read as the open question on this canvas rather than as a plan. A pivotal trial in post-traumatic stress disorder takes roughly three years from first participant to topline, so an approval decision before 2031 would require a start that has not happened yet.
There is a harder version of the question. The trial the FDA wants would need to solve the unblinding problem that no MDMA study has ever solved, in a population that is unusually likely to have taken the drug before. Nobody has yet published a design that does this. Until someone does, the right description of MDMA-assisted therapy in the United States is not delayed. It is stalled.
2028-06
plannedenterpriseProbability-weighted approval scenarios
Resubmission timing, approval probability, and how the R-MDMA programmes are priced against a stalled lead asset.
Everything else on this canvas is public record: registry entries, published papers, regulatory correspondence and company statements.
Blossom Enterprise models resubmission timing and approval probability for the MDMA-AT programme, and the comparator question this timeline keeps circling, which is whether a patented single enantiomer without a therapy protocol reaches a regulator before the racemic compound gets a second hearing. It is the one marker here you cannot verify from a registry, and it is deliberately the only thing behind a paywall.
MDMA
Midomafetamine: the research journey