How to Read a Psychedelic Clinical Trial: Blinding, Expectancy, Support and Follow-Up

Why these trials are hard to read

Headlines about psychedelic trials often say "breakthrough" or "failed". Neither word tells you much. Psychedelic trials have problems that most medicine trials do not: the drug is obvious to the person taking it, therapy is bundled with it, and enthusiasm in the field is high. This guide gives you questions to ask. It does not tell you what to take, and nothing here is a treatment recommendation. Evidence labels: promising but limited, mixed, insufficient evidence, Proper evidence not found.

1. What kind of question was the trial asking?

• Feasibility. Can the treatment be delivered and do people accept it? The anorexia nervosa psilocybin study was designed this way, with safety, tolerability and feasibility as its primary aim. [1]
• Safety. What side effects occurred, in whom, and for how long? The 2014 smoking pilot said an open-label design could not support firm conclusions about efficacy. [2]
• Efficacy. Does it work under controlled conditions against a comparison? The 2026 smoking trial compared psilocybin with the nicotine patch but described itself as a pilot and was unblinded. [3]
• Real-world effectiveness. Does it work in ordinary clinics for broader patients? The trials we read screened participants, for example enrolling only psychiatrically healthy smokers. [3] Proper evidence not found on real-world effectiveness in the sources we read.

A result at one level does not carry up to the next.

2. Blinding: could people tell what they got?

In a blinded trial, participants and raters should not know who received the active drug. With strongly psychoactive drugs they often guess correctly, called functional unblinding. [4] A 2026 systematic review of 112 randomized psychedelic trials found only 29.5% assessed blinding integrity, while 57.1% cited blinding as a limitation. [4] Psilocybin, LSD and ayahuasca studies frequently reported blinding failure above 90%, and no control strategy consistently achieved ideal blinding. [4]

Active placebos try to help. A migraine trial used diphenhydramine, which only partly mimicked psilocybin's acute effects. [5] An OCD trial used lorazepam. [6] Ask: did the paper test whether people guessed their group?

3. Expectancy: hope shapes results

Expectancy means what people expect to feel. Researchers note that as doses rise, hiding the treatment condition gets harder, and that expectancy can be measured and managed, not just feared. [7] A 2026 meta-analysis compared psychedelic therapy with antidepressants taken openly, which levels the blinding problem: psychedelic therapy (8 trials, 249 patients) was no more effective than open-label antidepressants. [8] Blinding changed results for antidepressants but not for psychedelic therapy, which the authors read as showing psychedelic trials are effectively open-label. [8]

4. Support and therapy: what is being tested?

Psychedelic-assisted therapy usually means preparatory sessions, a drug session, and integration sessions afterward. [9] If the drug and the therapy are tested together, you cannot tell which did the work. A 2025 review recommended factorial designs that test drug and psychotherapy separately. [10] See what is psychedelic-assisted therapy.

5. Follow-up: how long, and how complete?

Short follow-up can miss fading effects. In the OCD pilot, benefits diminished at 6 months but stayed substantial. [6] The Cochrane review of psychedelic therapy in life-threatening illness included studies lasting 6 to 12 months and rated much of the evidence low-certainty. [9] Also check dropout: in the 2026 smoking trial, 68 of 82 completed the 6-month check. [3]

6. Is it a result, a plan or one paper from a bigger study?

• Protocol. A 2025 paper described an open-label OCD pilot plan; it has no results. [11] A registered trial is not a positive result either.
• Repeat papers. The 2014 and 2017 smoking papers report the same 15 people. [2][12] The 2022 and 2024 cluster headache papers come from one trial and its extension. [13][14]
• Delay. The 2026 smoking trial collected data from 2015 to 2023. [3]
• Preprints are reports not yet peer reviewed. See reading early-stage research.

7. Who paid and who is involved?

Check funding and disclosures. The 2026 smoking trial authors report grants and fees from several psychedelic-related companies. [3] The Cochrane review noted that funding came from organizations that promote psychedelic-assisted therapy. [9]

Regulators

The US Food and Drug Administration finalized guidance for sponsors on psychedelic drug clinical investigations in July 2026. [15] We read only the web page, not the guidance text, so we do not summarize its content.

A quick checklist

1. Which substance, which condition, how many people?
2. Control group or open-label? Randomized?
3. Blinding tested? Expectancy measured?
4. Therapy included? Support described?
5. How long was follow-up, and who dropped out?
6. Is the outcome a clinical measure, not just a brain scan or biomarker?
7. Results, protocol or preprint?

Related: Psychedelic Clinical Trials and Mental Health Research, context and expectancy, finding registered trials.