The Psychedelic Renaissance: What the Trials Actually Show

The Controlled Substances Act of 1970 classified psychedelics as Schedule I substances, and nearly all legitimate psychedelic research in the United States halted for decades. That is the gap the word is pointing at, and our own research file frames the current wave of work explicitly as a reversal of that multi-decade research prohibition. It is a useful frame as long as it is not read as a story about a wronged science that was simply switched back on.

Because prohibition was not the only thing that killed the first wave. Psychedelic therapy research in the 1950s and 1960s also collapsed under poor methodology, inadequate safety protocols, and unethical practices in some of that era's studies. Our research file makes the point against itself: the current renaissance has to avoid repeating those failures. Hold on to that sentence. It reads very differently after 2024, when a modern regulator turned down a modern application over concerns from exactly this family of problems.

In 2006 Roland Griffiths and colleagues published the first rigorous double-blind psilocybin study in nearly forty years, in the journal Psychopharmacology. A high dose, 30 mg per 70 kg of body weight, produced mystical-type experiences in 72 percent of participants, and at a fourteen-month follow-up 67 percent rated the session among the five most personally meaningful experiences of their lives. Note carefully what that paper did and did not establish. It reported the character of an experience and how much it still mattered more than a year later. It did not report a clinical outcome in a patient population. Its importance is that it was rigorous, published, and possible at all, which reopened a door rather than proving anything through it.

The institution came next. Johns Hopkins formally established its Center for Psychedelic and Consciousness Research in 2019, building on a research program that had launched in 2000. Roland Griffiths, who lived from 1946 to 2023, led the Hopkins program until his death.

A randomized controlled trial run at Johns Hopkins in 2020, published as Davis and colleagues in JAMA Psychiatry in 2021, tested psilocybin-assisted therapy against major depressive disorder in 24 participants. It found rapid and sustained symptom reduction: 71 percent of participants showed a greater than 50 percent reduction in GRID-HAMD scores at four weeks, maintained at twelve months. The reported effect size, a Cohen's d of 2.3, substantially exceeded that of conventional antidepressants. Both halves of that description are load-bearing. An effect that size is a genuine finding and would be an extraordinary one if it holds. Twenty-four people is a small trial, and small trials carrying unusually large effects are precisely the ones that most often shrink when the work is repeated at scale.

Larger trials have since arrived, in a related indication. COMPASS Pathways' synthetic psilocybin formulation, COMP360, at a 25 mg dose, met its primary endpoint in two consecutive Phase 3 trials for treatment-resistant depression. COMP005 dosed 258 participants and reported topline results in 2025. COMP006 dosed roughly 581 and announced its primary endpoint on February 17, 2026: a statistically significant reduction on the MADRS depression scale against a 1 mg comparator dose, at p less than 0.001. In COMP006, 39 percent of the participants receiving 25 mg reached a clinically meaningful response, defined as an improvement of 25 percent or more on MADRS, against 25 percent reported for COMP005. A 26-week durability readout from COMP006's Part B followed on July 7, 2026, reporting treatment effects sustained after one or two doses.

Those are large trials and a real evidentiary advance over the small studies that defined the early 2020s. They are also, as things stand, company-announced topline statistics rather than complete peer-reviewed publications, and psilocybin is not approved for anything. A large positive Phase 3 result and an approved medicine are two different facts. This article keeps them apart everywhere, and three sections from here is the reason why.

An open-label pilot study of psilocybin for smoking cessation at Johns Hopkins, led by Matthew Johnson and published in 2014, found 80 percent biologically confirmed abstinence at six months. The figure usually quoted beside it is roughly 35 percent for varenicline, the standard comparator drug. The number that decides how much weight this can carry is neither of those: the sample was fifteen people, and the study was open-label, which means nobody involved was blinded to anything. It is a striking pilot result, and a pilot result is a reason to run the real trial rather than a substitute for having run it.

Alcohol has a stronger study behind it. A randomized controlled trial published in JAMA Psychiatry in 2022, by Bogenschutz and colleagues, gave psilocybin-assisted therapy to 93 participants with alcohol use disorder and compared it against an active placebo, diphenhydramine, with matched therapy on both sides. Heavy drinking days fell by 83 percent at eight months in the psilocybin group, against 51 percent in the active-placebo group. Note the active placebo, and note that both groups got the therapy. That design is the field's main defense against a problem this article comes to shortly, and it is only a partial defense.

MDMA, or 3,4-methylenedioxymethamphetamine, was first synthesized in 1912 by the Merck chemist Anton Köllisch, and patented as a chemical precursor compound rather than as a psychoactive drug. It was rediscovered for possible therapeutic use by the chemist Alexander Shulgin in the 1970s, and an estimated 500,000 therapists and clients had used it before it was classified as Schedule I in 1985. Psychedelic research did not have one shutdown. It had two, fifteen years apart.

Its pharmacology is also distinct from psilocybin's, which matters to what it is being tested for. MDMA's core action is a massive presynaptic release of serotonin, roughly 80 percent, dopamine, roughly 15 percent, and norepinephrine, roughly 5 percent, together with increased oxytocin secretion. That is a different pathway from psilocybin's serotonin-receptor agonism and its effects on the default mode network. MDMA is classed as an entactogen, a category named for the feelings of emotional closeness and trust it produces, rather than as a classic serotonergic psychedelic.

The Phase 3 program built on that. The first Phase 3 trial of MDMA-assisted therapy for PTSD, run by MAPS and published as Mitchell and colleagues in Nature Medicine in 2021, known as MAPP1, randomized 90 participants after a psychiatric medication washout to either MDMA with therapy or placebo with the same therapy. After three MDMA sessions plus preparatory and integration psychotherapy, 67 percent of the MDMA group no longer met the diagnostic criteria for PTSD, against 32 percent in the placebo-with-therapy group.

A second Phase 3 trial followed. MAPP2, published by Mitchell and colleagues in Nature Medicine in September 2023, also met its primary endpoint, and it enrolled a more diverse population than MAPP1: people of color made up roughly half of its sample. So both Phase 3 trials of MDMA-assisted therapy were positive on efficacy. That is the fact which makes what happened next difficult, and it is the fact most often dropped from both of the popular versions of this story.

One structural detail is needed before the rejection, because it decides who was rejected. MAPS PBC, the public benefit corporation MAPS founded in 2014 to carry MDMA-assisted therapy through FDA approval, was renamed Lykos Therapeutics in January 2024 following a 100 million dollar Series A financing round. MAPS keeps governance ties, holding 10-to-1 voting power and appointing six of Lykos' eight board seats, but Lykos operates as a legally separate for-profit entity from the MAPS nonprofit. Lykos is the entity the FDA's letter was formally issued to.

On June 4, 2024, an eleven-member FDA Psychopharmacologic Advisory Committee voted 9 to 2 that the available data do not show MDMA-assisted therapy's effectiveness for PTSD, and 10 to 1 that its benefits do not outweigh its risks. The committee's stated concerns were specific rather than atmospheric: functional unblinding, meaning that participants could tell whether they had received MDMA; the particular form of psychotherapy the drug was paired with; and ethical concerns about documented misconduct in some MAPS-sponsored trials.

On August 9, 2024, the FDA issued a Complete Response Letter rejecting Lykos Therapeutics' New Drug Application for MDMA-assisted therapy for PTSD. The agency stated that the application could not be approved based on data submitted to date, and requested an entirely new Phase 3 trial. As of July 2026, MDMA remains federally Schedule I, with no FDA-approved use outside clinical trials and FDA Expanded Access, and no public resubmission date has been announced.

Two readings of that letter circulate, and this article refuses both. The first calls it a technicality, a piece of bureaucratic obstruction standing between patients and a medicine that works. It was not. An eleven-member advisory committee voted 9 to 2 against effectiveness and 10 to 1 against a favorable balance of benefit and risk, and the agency then demanded an entirely new Phase 3 trial. That is a serious scientific and regulatory setback for the whole field, not a footnote and not a paperwork problem. The second reading calls it proof that MDMA does not work. It was not that either: both Phase 3 trials met their primary endpoints, and the agency's stated objection was to how that evidence was generated, not to the direction it pointed. Refusing one of these framings does not commit you to the other. And the field's own history, the one in section 01, says concerns about trial conduct in this particular research tradition are the kind worth taking seriously.

A psychedelic's subjective effects are usually unmistakable to the person having them. That makes a genuinely blind placebo-controlled trial very difficult to run, because participants, and often the clinicians observing them, can tell who received the drug. The field calls this functional unblinding, and it is not a marginal critique from outside the tent: it was explicitly one of the FDA advisory committee's reasons for its 2024 vote, and Muthukumaraswamy and colleagues argued in 2021 that published effect sizes across the field may be inflated by expectancy bias. If people know they received the real thing, and they entered the trial hoping it would work, some unknown portion of the result is that hope. Active placebos are the partial answer: niacin, a low dose of psilocybin, or diphenhydramine, as in the alcohol trial above. They narrow the gap. They do not close it, and the researchers using them do not claim otherwise.

Neuroimaging studies using fMRI during psilocybin sessions show decreased activity and decreased connectivity in the default mode network, the brain system associated with self-referential thought. Robin Carhart-Harris, at Imperial College London, proposed that this disruption is what mediates psilocybin's ego-dissolution effects and its therapeutic ones, in work published with colleagues in PNAS in 2012. The imaging finding is solid. The claim that the disruption is the mechanism of a clinical benefit is a hypothesis attached to that finding, and the two deserve different amounts of confidence.

The broader version is the entropic brain hypothesis, put forward by Carhart-Harris in 2014: that psychedelics increase neural entropy, meaning signal diversity, temporarily destabilizing rigid and maladaptive patterns of brain activity and allowing a therapeutic reorganization to follow. It is consistent with measured increases in neural-complexity measures under psilocybin. The mechanistic details remain genuinely debated in the literature, and this article carries that debate open rather than choosing the tidiest available version of it.

There is one further finding here that is easy to misuse, so it is stated narrowly. Multiple studies support what the field calls the mystical experience hypothesis: that therapeutic outcomes correlate with the intensity of the mystical-type experience during a session rather than with the drug's pharmacological action alone. Griffiths and colleagues found in 2008 that mystical-experience scores predicted long-term positive changes in attitudes, mood and behavior. Used precisely, this is a clinical-outcome predictor and nothing more: a score on a questionnaire that turns out to forecast who improves.

Used imprecisely, it becomes something else. Some researchers argue that psychedelic experiences deliver genuine metaphysical insight, the position usually filed under perennial philosophy. That claim cannot be evaluated by clinical trial methodology at all, which places it outside medical research and outside this file. This library carries it elsewhere, in the article on ego dissolution in The Doors of Perception wing. Here the experience is a variable in a study, and it stays one.

On March 5, 2019, the FDA approved SPRAVATO, esketamine, a nasal spray for treatment-resistant depression in adults, to be used in conjunction with an oral antidepressant. It was the first approved medication for depression with a genuinely novel mechanism of action, antagonism at the NMDA receptor, since fluoxetine in 1987. It is dispensed and administered only at clinics certified under a Risk Evaluation and Mitigation Strategy, because of the risk of sedation and dissociation and because the drug carries abuse potential.

Esketamine is not the same thing as the ketamine given at ketamine clinics, and the two get merged constantly in ordinary conversation. Esketamine is the S-enantiomer, one of the two mirror-image forms of the molecule, delivered as an FDA-approved nasal spray inside that REMS framework. The racemic ketamine used off-label at infusion clinics for depression, anxiety and other conditions is a 50/50 mixture of both mirror images, typically given intravenously. Racemic ketamine is FDA-approved as an anesthetic and for no psychiatric indication at all. Its use for depression is off-label, generally not covered by insurance, and outside the specific REMS framework that governs esketamine.

The FDA granted Breakthrough Therapy designation to COMPASS Pathways' psilocybin for treatment-resistant depression in October 2018, and to MAPS' MDMA-assisted therapy for PTSD in August 2017. The designation is granted when preliminary clinical evidence indicates a drug may show substantial improvement over an existing therapy on a clinically significant endpoint, and it expedites the agency's development and review process. It is not a marketing authorization and it is not an approval. The proof of that is what became of these two designations: one was rejected in 2024, and the other is still pending.

On April 18, 2026, President Trump signed an executive order titled Accelerating Medical Treatments for Serious Mental Illness, directing the FDA and the DEA to establish accelerated pathways for psychedelic drug development and review. It included Commissioner's National Priority Vouchers for drugs holding Breakthrough Therapy designation, which could shorten review periods from months to weeks, and allocated 50 million dollars for federal and state collaboration on serious mental illness research.

Six days later, on April 24, 2026, FDA Commissioner Marty Makary announced that the agency had issued three Commissioner's National Priority Vouchers under the new pathway: to COMPASS Pathways' psilocybin, COMP360, for treatment-resistant depression; to the Usona Institute's psilocybin for major depressive disorder; and to Transcend Therapeutics' methylone, TSND-201, for PTSD.

Two states got there first, by a route that has nothing to do with the FDA. Oregon's Measure 109, passed in 2020, created a state-regulated pathway in which adults aged 21 and over may use psilocybin at licensed service centers under supervision, with no medical prescription and no diagnosis required. Colorado's Proposition 122 decriminalized personal psilocybin possession statewide and established licensed healing centers beginning in 2025. Neither state has legalized MDMA. These are state-level regulated-access models, legally and procedurally distinct from federal drug approval, and neither one is evidence about the other.

The right-hand column of that table exists to refuse one specific move, made constantly in coverage of this subject. A Breakthrough Therapy designation is not an approval. A positive Phase 3 topline announcement is not an approval. A submitted or rolling New Drug Application is not an approval. A priority voucher is not an approval. As of July 2026 the only FDA-approved compound in this article is esketamine, and no favorable trial result and no favorable regulatory wind changes that. The next time a headline says psychedelics have been approved, the question that settles it is narrow: approved for what indication, by which agency, on what date.

The first is that psychedelics are safe. They are not universally safe, and the trial protocols themselves are the evidence. MDMA carries real cardiac contraindications, through serotonergic cardiotoxicity. In people with a personal or family history of psychotic disorders there is a risk of prolonged psychosis. Cases of Hallucinogen Persisting Perception Disorder, in which perceptual disturbance persists after the drug has gone, are rare and documented. These substances are studied with extensive screening and with therapist support present throughout precisely because they are not safe for everyone, and that screening is not a formality that a supervised setting could dispense with.

The second refusal runs the other way, against an old fear rather than a new hope, and honesty requires carrying it too. The claim that psilocybin or LSD cause chromosome damage, brain lesions, or persistent psychosis in healthy populations came out of methodologically flawed studies from the 1960s and 1970s. Modern controlled published evidence demonstrates no lasting cognitive impairment and no genetic damage at clinical doses, as Johnson and colleagues set out in Neuropharmacology in 2018. Read the qualifiers, because they are what make the two refusals compatible: healthy populations, and clinical doses. The old scare literature was wrong. The screening criteria in the paragraph above are still there for a reason.

The third is the one this article works hardest against, because it is the one the coverage keeps manufacturing. These are not miracle cures, and nothing here should be described as a cure for depression, for PTSD, for addiction, or for anything else. Every efficacy result above arrives attached to a sample size, a comparison group and a methodological caveat, and a cure framing erases all three at once. There is no approved psychedelic-assisted therapy for PTSD, for major depressive disorder or for alcohol use disorder. The one approved compound in the neighborhood, esketamine, is approved as an addition to an oral antidepressant for treatment-resistant depression specifically, which is to say for people whom other treatments have already failed. That is a real and useful thing to have. It is not a cure, and it is not shaped like one.

Therapeutic psilocybin requires six to eight hours of therapist support for every dosing session, on top of intensive screening and separate preparation and integration sessions. An estimated cost of 10,000 to 25,000 dollars per treatment course follows from that structure. Without insurance coverage or policy reform, access may be limited to affluent patients. This is an access and equity concern distinct from the science, and it is worth naming separately, because a treatment that works and cannot be delivered at scale is a different kind of success from the one the headlines imply.

Durability is unanswered. Most published follow-up data extends only twelve to eighteen months, so the long-term durability of a single course or a few sessions, at ten years and beyond, is simply unknown. Whether periodic booster sessions would be needed to hold a benefit is an open clinical question with no current answer. The 26-week readout from COMP006, useful as it is, does not touch that question.

One frequently trailed application has no human evidence behind it at all. The idea that psychedelic therapy might address neurodegenerative conditions such as Alzheimer's or Parkinson's through enhanced neuroplasticity is an emerging hypothesis with a preclinical basis: Ly and colleagues showed in 2018 that psilocybin promotes dendritic spine growth and BDNF release in mice. Human clinical evidence for neurodegenerative applications is entirely absent. Mice are not a small trial. They are not a trial.

Which leaves the question the whole field is currently waiting on. COMPASS Pathways is targeting completion of a rolling New Drug Application in the fourth quarter of 2026, with a possible FDA decision in late 2026 or early 2027. An approval would be the first the FDA has ever granted to a classic psychedelic. Whether it will come is genuinely open rather than a formality, and the reason it is open sits two years behind it in this same article: MDMA-assisted therapy, from a closely related drug class, also cleared two positive Phase 3 trials and was rejected anyway. Positive Phase 3 data alone did not guarantee approval for MDMA, and there is no certainty it will prove sufficient for psilocybin.