Psilocybin: what the evidence shows

Two Phase 3 trials of COMP360 psilocybin met their main goal. In the first (258 people), one 25 mg dose beat placebo by 3.6 points on the MADRS depression scale at week 6; in the second (581 people), 25 mg beat a 1 mg dose by 3.8 points.¹²

Those are modest average differences, and individual results vary a lot. In the second trial, 39% of people on 25 mg saw their symptoms drop by at least a quarter by week 6, and those responses held to at least six months.³

In the first trial's open-label extension, about 80% of the 25 mg group took another dose; at one year their depression scores averaged about 13 points below where they started.⁴

The FDA agreed to a rolling review and gave COMP360 a national priority voucher, which can shorten review to one or two months once the application is complete. Compass expects to finish filing in Q4 2026.⁵⁶³

This builds on a 233-person Phase 2b trial published in the New England Journal of Medicine.⁷

Not every trial has been positive. In EPISODE, an academic Phase 2b trial in Germany (144 people), 25 mg didn't significantly beat an active placebo on its main measure: 17% responded, versus 11%. Thoughts of suicide on dosing days were slightly more common with psilocybin.⁸

In Australia, authorised psychiatrists have been allowed to prescribe psilocybin for treatment-resistant depression since 2023.⁹

In a Phase 2 trial of 104 adults with major depression, one dose of psilocybin cut depression scores by about 12 points more than an active placebo at six weeks.¹⁰

An earlier Johns Hopkins trial (27 people) compared two psilocybin sessions with a waiting list: 71% responded at four weeks. A waiting list isn't a placebo, so this overstates the drug's own effect.¹¹

Head-to-head against the antidepressant escitalopram, psilocybin was not significantly better on the main measure, though some secondary results favoured it.¹²

Usona Institute is running a Phase 3 trial (uAspire). The FDA gave this programme a national priority voucher in April 2026. No Phase 3 results have been published.¹³⁶

A related molecule, HLP003 (Helus Pharma), finished enrolling its first Phase 3 trial in 2026, with results expected in Q4 2026. It isn't psilocybin itself.¹⁴

In a double-blind trial of 93 adults with alcohol dependence, psilocybin plus therapy roughly halved the share of heavy-drinking days compared with an active placebo over 32 weeks.¹⁵

In a randomised, placebo-controlled Phase 2 trial of 73 adults with GAD, two psilocybin sessions cut anxiety scores by 12.8 points versus 3.6 with placebo. These are company-reported results; the full paper isn't out yet.¹⁶

Two randomised trials in people with life-threatening cancer found large, lasting drops in depression and anxiety after one supported session.¹⁷¹⁸

A newer open-label study of 30 people with cancer and depression tested psilocybin in a group format and also saw sizeable improvements.¹⁹

In a randomised pilot of 82 daily smokers, about 40% of the psilocybin group had stayed off cigarettes at six months, versus about 10% on nicotine patches. Everyone knew which treatment they got.²⁰

This follows an earlier 15-person open-label pilot that reported high quit rates.²¹!

In an open-label pilot of 5 adults with fibromyalgia, two psilocybin sessions with therapy were followed by less pain and better sleep a month later, with no serious side effects. Recruitment stopped early, and there was no comparison group.²²

In an open-label study of 10 women with anorexia, psilocybin was safe and tolerable; four had meaningful drops in eating-disorder symptoms at three months.²³

A 2026 pilot with 21 women saw eating-disorder scores fall at every follow-up, but body weight didn't change meaningfully, and there was no comparison group.²⁴

In a 10-person placebo-controlled study, one dose reduced migraine days over the following two weeks.²⁵

In cluster headache, a small trial didn't show a clear drop in attack frequency, though a later blinded extension phase did.²⁶²⁷

In a randomised trial of 15 adults, weekly high doses of psilocybin reduced OCD symptoms more than an active placebo over four weeks.²⁸

In a UK study of 19 adults, a 10 mg dose eased compulsions more than a 1 mg dose for about a week. Doses were given in a fixed order rather than at random, and the effect faded within a month.²⁹

That team's first study, with 9 patients, dates back to 2006.³⁰!

In an open-label Phase 2 study of 22 people with PTSD, one 25 mg dose was followed by large symptom drops at 12 weeks, with no serious side effects. Without a placebo group, this can't tell us how much was the drug.³¹