Research desk · Psychedelics
Clients are asking about psilocybin. Some are planning a trip to Oregon or Colorado, and some already went. The 2025 and 2026 research is more mixed, and more interesting, than either the hype or the backlash. Here is a clinician’s read.
Psilocybin occupies a strange legal and clinical space in late 2026. It is still a Schedule I drug federally, and no psilocybin product is FDA-approved. Compass Pathways says it is on track to complete a rolling new drug application for COMP360, its psilocybin formulation, in the fourth quarter of 2026. It anticipates a US launch in the first half of 2027, contingent on FDA approval and DEA rescheduling. Meanwhile, Oregon runs state-regulated psilocybin services, and Colorado licenses healing centers and facilitators through its Natural Medicine program.
So your clients have legal access in some states, years of enthusiastic coverage behind them, and a pharmaceutical approval possibly on the way. What they rarely have is an accurate picture of the evidence. Neither do many clinicians, partly because the media story and the trial story have drifted apart. A 2026 systematic review compared 70 psilocybin trials with 172 media articles. [11] Media coverage frequently cited PTSD (43% of articles) and anxiety (42%) as indications, even though trial evidence was absent for PTSD and limited for anxiety. Only 30% of media articles mentioned safety at all. In the trials themselves, long-term safety was assessed in just 6%.
The null trial. EPISODE, published in JAMA Psychiatry in 2026, randomized 144 adults with treatment-resistant depression to psilocybin 25 mg, psilocybin 5 mg, or nicotinamide as an active placebo, each with adjunct psychotherapy. [1] Response rates on the primary endpoint were 17.0%, 12.5%, and 10.6%, respectively. The first comparison (25 mg vs placebo) was not significant (adjusted OR 1.73, 95% CI 0.53–6.23), so formal testing stopped there. Secondary measures gave what the authors called exploratory evidence of a clinically meaningful effect. They also reported safety signals: more suicidal ideation on dosing days with 25 mg (4% vs 1–2%), and two serious adverse reactions, including one case of hallucinogen persisting perception disorder. The authors themselves called it an inconclusive trial.
The strong trial. A 2026 Nature Medicine feasibility RCT, run at one NHS site in England, randomized 60 adults with treatment-resistant depression to a single 25 mg dose of psilocybin or placebo, with preparation, dosing support, and integration. [2] At week 3, the adjusted MADRS difference was −10.41 points (Cohen’s d = −1.70) in favor of psilocybin, and it held at week 6. The psilocybin arm had more nonserious adverse events (164 vs 123). Its primary aims were feasibility outcomes, and the authors framed the result as support for a confirmatory trial.
The durability question. A Swedish trial of 35 adults with moderate to severe recurrent depression found a significant MADRS difference at day 8 (−7.27), day 15 (−11.03), and day 42 (−8.33). By day 365 the difference was no longer significant (−3.68). [3] Two people in the psilocybin group reported persistent, severe anxiety that required medical attention. Most other adverse events were transient and mild to moderate.
Beyond psilocybin. A small phase IIa trial of intravenous DMT, a short-acting psychedelic, reported a MADRS difference of −7.35 at two weeks in 34 adults with major depression. [4] The research field is moving toward shorter sessions, which could change the cost and logistics of care.
Many people in a psilocybin trial can tell within the first hour whether they got the drug. That isn’t a small technical flaw. It strikes at the heart of how we estimate effect sizes. A 2026 systematic review of 112 psychedelic RCTs found that only 29.5% even assessed blinding. [5] Where it was measured, psilocybin, LSD, and ayahuasca studies frequently reported blinding failure above 90% among both participants and raters. Inert-placebo MDMA trials exceeded 85%. No control strategy reliably fixed it.
Two other 2025–2026 analyses show what this does to the numbers. The first compared control arms across depression trials. [6] The active arms looked similar: standardized mean change of 1.21 for psilocybin, 1.28 for SSRIs, and 1.43 for esketamine. The control arms did not. Psilocybin control groups improved far less (0.50) than SSRI controls (1.00) or esketamine controls (1.12). In other words, part of psilocybin’s large advantage may come from disappointed control participants rather than an unusually strong drug effect.
The second meta-analysis took the logical next step. Psychedelic-assisted therapy is effectively open-label, so it compared that therapy with open-label conventional antidepressants instead of with blinded placebo. [7] Under those equal-unblinding conditions, psychedelic-assisted therapy was no more effective. The estimated difference was 0.3 Hamilton points, favoring the open-label antidepressants (95% CI −1.39 to 1.98).
None of this means psilocybin does nothing. It means the large effect sizes in popular coverage almost certainly overstate the drug’s specific contribution. Expectancy, setting, and the therapeutic relationship are doing some of the work. That’s familiar ground for therapists.
One of the more clinically useful 2026 findings is about the therapy around the drug. A meta-analysis of 12 controlled trials (733 participants) found a large overall effect of psychedelic-assisted therapy on depressive symptoms compared with controls (Hedges g = −0.84). [8] The authors then asked whether the amount of accompanying psychological work mattered. More hours of preparation were associated with greater symptom reduction (β = −0.13). Hours of post-dosing integration and total session count were not significantly associated. They also judged 75% of the trials to be at high risk of bias, so the finding is a lead, not a rule.
For ketamine, the picture is thinner still. A 2026 systematic review of ketamine-assisted psychotherapy found 72 studies but only 11 randomized trials. Only two of those randomized the psychotherapy component itself, and neither found that psychotherapy added benefit. [9] The authors concluded that current evidence does not support an added benefit or synergy, while stressing the need for better-controlled trials.
Trials screen participants carefully. State programs serve whoever walks in. That makes the first outcome data from Oregon’s regulated psilocybin services, published in JAMA Network Open in 2026, worth reading in full. [10] The cohort included 346 clients at 24 of the state’s 26 licensed service centers, supervised by 83 licensed facilitators, and retention was high (90.2% at three months).
Four people (1.2%), all psychedelic-naive, had adverse behavioral reactions requiring medical attention. Between 1.6% and 2.3% described the experience as harmful at follow-up points. At one month, 92.2% felt they had benefited. At three months, moderate-to-severe depression symptoms were less common (RR 0.40), as were anxiety and PTSD symptoms. Note who was in the sample: 86.7% White, 76.6% college-educated, and 29.0% with household income above $200,000. There was no control group, so this describes outcomes in people who chose and could afford the service. It does not show that psilocybin caused the change.
A way to open the conversation:
“Some people I work with have been curious about psilocybin or have tried it. If that’s ever on your mind, I’d rather hear about it than not. I won’t judge it, and I can share what the research actually shows.”
And when a client says the articles call it a cure:
“The best trials show real promise for depression, but the effects are smaller and less certain than the headlines suggest, and they don’t always last. Let’s look at what you’re hoping it will change, and what else could help with that.”
Each card links to the PubMed record. Numbers are reported the way the authors reported them. Read the abstract before you quote any of them.
Mertens LJ, Koslowski M, Betzler F et al.
EPISODE, 144 adults with treatment-resistant depression. Response 17.0% (25 mg) vs 12.5% (5 mg) vs 10.6% (nicotinamide); primary comparison not significant (aOR 1.73, 95% CI 0.53–6.23). Suicidal ideation on dosing days 4% vs 1–2%; two serious adverse reactions, including one case of HPPD.
Rucker JJ, Mantingh T, Kerr-Gaffney J et al.
60 adults with TRD at one NHS site. MADRS difference at week 3 −10.41 (95% CI −14.86 to −5.95; d = −1.70), sustained at week 6. 164 vs 123 nonserious adverse events.
Yngwe H, Plavén-Sigray P, Ekman CJ et al.
35 adults with recurrent MDD, psilocybin vs niacin. MADRS difference −7.27 at day 8 and −11.03 at day 15; not significant at day 365 (−3.68). Two participants had persistent severe anxiety requiring medical attention.
Erritzoe D, Barba T, Benway T et al.
IV DMT, 34 adults with MDD. MADRS difference −7.35 at 2 weeks (95% CI −13.62 to −1.08). No serious adverse events.
Orsini DK, Wong S, Di Luch S et al.
112 psychedelic RCTs. Only 29.5% assessed blinding integrity. Psilocybin, LSD, and ayahuasca studies frequently reported blinding failure above 90%; inert-placebo MDMA trials exceeded 85%.
Hieronymus F, López E, Werin Sjögren H et al.
17 depression trials. Active arms were similar (SMC 1.21 psilocybin, 1.28 SSRI, 1.43 esketamine), but psilocybin control arms improved less (0.50 vs 1.00 and 1.12).
Williams ZJ, Barnett H, Szigeti B
Psychedelic-assisted therapy vs open-label antidepressants. No more effective: difference 0.3 HAM-D points favoring open-label antidepressants (95% CI −1.39 to 1.98).
Florineth GA, Klima I, Boeker AL et al.
12 controlled trials, 733 participants. Overall g = −0.84. More preparation hours associated with greater improvement (β = −0.13); integration hours not significant. 75% of trials at high risk of bias.
Veraart JKE, Schimmers N, Breeksema JJ et al.
Ketamine-assisted psychotherapy: 72 studies, 11 RCTs. Only two randomized the psychotherapy component; neither found added benefit.
Korthuis PT, Cook RR, Gregoire D et al.
346 clients of Oregon psilocybin services. 1.2% had adverse behavioral reactions needing medical attention (all psychedelic-naive); 92.2% felt they benefited at 1 month; moderate-to-severe depression RR 0.40 at 3 months. The sample was 86.7% White, and 29.0% had household income above $200k.
Madden K, Flood B, Kashir I et al.
70 psilocybin trials vs 172 media articles. Media cited PTSD (43%) and anxiety (42%) despite absent or limited evidence. Only 30% of articles mentioned safety; long-term safety was assessed in 6% of trials.
Education for clinicians, not medical or legal advice. This page summarizes published research for licensed mental health professionals. It is not a treatment recommendation for any individual, and it does not replace your own clinical judgment, your licensing board’s rules, or consultation with a prescriber or attorney. Evidence reviewed through October 24, 2026. Spot an error? Tell us and we will fix it.
Keep reading