Psilocybin for depression.
What the trials actually show.

The four most consequential trials

1. Carhart-Harris et al., Imperial College (NEJM, 2021)

Fifty-nine patients with moderate-to-severe depression were randomised to either two psilocybin sessions (25mg, 3 weeks apart) plus daily placebo, or six weeks of daily escitalopram (an SSRI) plus two low-dose placebo sessions. The primary endpoint — change on a specific depression rating scale — was statistically equivalent between groups. Secondary measures, including remission rate and broader symptom measures, favoured psilocybin.

What the trial demonstrated: two psilocybin sessions over six weeks produced antidepressant effects comparable to six weeks of standard SSRI treatment. What it did not demonstrate: psilocybin superiority. The trial was widely misreported as showing psilocybin "beat" SSRIs; it didn't. It showed parity, which is itself notable.

2. Goodwin et al., COMPASS Pathways Phase IIb (NEJM, 2022)

233 patients with treatment-resistant depression randomised to one of three doses of synthetic psilocybin (25mg, 10mg, or 1mg as active placebo). At three weeks, the 25mg arm showed a response rate of 37% and a remission rate of 29%. The 10mg arm was statistically indistinguishable from the placebo dose.

What the trial demonstrated: a dose-response relationship, with meaningful efficacy at 25mg in patients who had failed multiple prior antidepressants. The response rate at week 1 was 67%; by week 12, it had decayed to 25%. Durability without repeat dosing is a real question.

3. Gukasyan et al., Johns Hopkins long-term follow-up (J Psychopharmacology, 2022)

27 patients from a previous psilocybin trial for major depressive disorder were followed up at 12 months. 75% maintained clinically significant response and 58% met criteria for full remission. Effect sizes at 12 months were comparable to those at 1 month post-treatment.

What the trial demonstrated: durability is possible. This is a smaller, more selective sample than COMPASS, and the patient population was less treatment-resistant. But the durability finding is significant — most antidepressants require continuous dosing.

4. Griffiths et al., end-of-life anxiety in cancer (J Psychopharmacology, 2016)

51 patients with advanced cancer and clinically significant anxiety or depression received either a single high-dose or low-dose psilocybin session. At six months, 80% of the high-dose group showed clinically significant reductions in both anxiety and depression measures.

This was the first major modern trial to demonstrate durable benefit from a single psychedelic session, and it remains one of the most clinically striking results in the field.

What the data does and doesn't support

It does support: psilocybin produces rapid, clinically significant antidepressant effects in adults with both depression and treatment-resistant depression. Effects can be durable for months to a year, particularly in less treatment-resistant patients. The therapeutic window is broad — most patients tolerate it well. There is a meaningful dose-response.

It does not (yet) support: psilocybin's superiority over standard treatments — the comparative evidence shows roughly parity with SSRIs. Universal applicability — the trials select for psychiatric stability, and patients with active psychosis, bipolar disorder, or severe personality disorders are typically excluded. Effects without the psychotherapy wraparound — every trial pairs the drug with therapy preparation, support during, and integration after. The drug-only effect has not been isolated.

The methodological caveats

Blinding psychedelic trials is hard. A 25mg psilocybin dose is unmistakeable; patients know which arm they're in. Researchers attempt to control for this with active placebos and with statistical analyses, but the inability to truly blind remains a real critique.

Patient self-selection matters. People who volunteer for psychedelic trials tend to have positive expectations. Expectancy effects in this kind of intervention are well-documented and probably substantial.

Sample sizes are still relatively modest. The largest single-trial sample to date is COMPASS at 233 patients. By pharmaceutical-trial standards this is small, particularly given the heterogeneity of depression as a diagnosis.

What's next

COMPASS Pathways' Phase III trial in treatment-resistant depression is ongoing, with first results expected in 2026. Usona Institute is conducting Phase III work in major depressive disorder. Both trials are larger, more rigorous, and structured for FDA submission. If either reads out positive, regulatory approval in the US is plausible within 2–3 years.

Outside the US, regulatory landscape is shifting. Australia approved psilocybin for treatment-resistant depression in 2023. Canada has a special-access pathway. Several US states (Oregon, Colorado) have legalised regulated therapy.

Practical implications

For someone considering psilocybin therapy for depression, the evidence is real but not yet decisive. If you are in active treatment-resistant depression, the response rates in the trials are meaningful, particularly compared to the modest effect sizes of fourth- and fifth-line conventional treatments.

But "psilocybin works for depression in a clinical trial setting with extensive psychotherapeutic support" does not automatically translate to "a psilocybin retreat will work for me." The trial conditions — high-quality therapeutic relationship, structured preparation, individualised integration — are not always replicated in retreat settings. The provider matters. See verified psilocybin providers.

For the full set of cited statistics with primary-source links, see our statistics page.

Last updated May 2026.