Why can't we trust the results from psychedelic trials?
The core issue is that psychedelics produce such distinctive subjective effects—hallucinations, altered perception, ego dissolution—that participants and even therapists can usually tell who got the drug and who got the placebo. This is called 'unblinding,' and it breaks the fundamental assumption of a placebo-controlled trial: that both groups believe they might be receiving the real treatment. Two of the papers here conducted systematic reviews of psychedelic randomized controlled trials (RCTs) and found that most did not even report whether blinding was successful [2]. The authors argue that because of unblinding, treatment effect sizes in these trials are 'likely over-estimated' [2]. Another paper makes the same point more bluntly: psychedelic trials 'are likely to significantly overestimate treatment effects by design due to unblinding and expectancy effects' [1].
This isn't just a theoretical worry. One study used computational modeling to show how weak blinding, combined with positive expectations, can create what they call 'activated expectancy bias' (AEB) [4]. When they re-analyzed data from a real 'self-blinding' microdosing trial, they found that the apparent benefits of microdosing over placebo could be entirely explained by AEB—meaning the results were at risk of being false positives [4]. The authors conclude that microdosing may be best understood as an 'active placebo,' where the act of taking something you believe will help you produces the effect, not the drug itself [4].
Does this mean psychedelics don't work?
No—and the researchers themselves are careful not to throw the baby out with the bathwater. The same paper that warns about overestimation also notes that 'problems in blinding should not automatically disqualify medications from licensing decisions' [1]. They point out that many other medical treatments—like surgery, psychotherapy, and some pain medications—face similar blinding challenges and are still considered effective [1]. The real lesson is that we need to interpret the current evidence with caution, not dismiss it entirely.
In fact, one paper argues that the powerful placebo effects in psychedelic therapy are not a bug but a feature. They propose a framework inspired by the psychedelic concept of 'set and setting'—meaning the drug's effects are shaped by the person's mindset and the physical/social environment [3]. Instead of trying to eliminate these 'nonspecific' factors, they suggest we should harness them ethically to improve clinical outcomes [3]. This reframes the debate: the question isn't whether expectancy is contaminating the results, but how we can design treatments that make the most of both the drug and the context in which it's given.
Another paper focuses specifically on microdosing—taking tiny, sub-hallucinogenic doses—and notes that the claimed benefits are 'likely' driven by expectations and confirmation bias, with 'little supporting empirical research' [5]. But this doesn't apply equally to full-dose psychedelic therapy, where the subjective experience is so intense that it's hard to attribute all the benefits to mere expectation. The evidence here suggests that while expectancy inflates effect sizes, it probably doesn't account for all of them—especially in high-dose, therapist-supported sessions.
How can future research fix this?
The papers agree on several practical fixes. First, trials should routinely measure how well blinding worked—asking participants at the end which group they think they were in—and report those results [2]. Second, researchers can use 'active placebos' (a substance that produces some mild side effects but not the psychedelic experience) to make blinding more convincing [1]. Third, using independent raters who don't know which treatment a participant received can reduce bias in outcome assessments [1].
One paper introduces a new statistical tool called the 'Correct Guess Rate Curve' (CGRC) that can estimate what the results of a perfectly blinded trial would have been, based on data from an imperfectly blinded one [4]. They recommend this approach to correct for expectancy bias in future studies. Another paper suggests supplementing RCTs with naturalistic studies—observing people who use psychedelics in real-world settings—to see if the benefits hold up outside the highly controlled (and expectation-laden) trial environment [1]. Together, these recommendations show that the field is aware of the problem and is actively developing solutions, rather than ignoring it.
About These Sources
This answer is built on 5 peer-reviewed studies — published from 2021 to 2023, 3 in Q1 journals, collectively cited 426 times — selected as the most relevant from 5 studies that passed quality screening, drawn from 45 papers retrieved from a database of over 500 million.
Sources used in this answer
Expectancy in placebo-controlled trials of psychedelics: if so, so what?
Argues that psychedelic trials likely overestimate treatment effects due to unblinding and expectancy, but this is not unique to psychedelics; suggests practical mitigations like active placebos and independent raters, and notes that imperfect blinding should not automatically disqualify a treatment from licensing.
Blinding and expectancy confounds in psychedelic randomized controlled trials
Conducts a systematic review of psychedelic RCTs (1990–2020) and finds that most did not report pre-trial expectancy or blinding success; concludes that effect sizes are likely overestimated due to de-blinding and response expectancy, and urges caution in interpreting results.
Harnessing placebo: Lessons from psychedelic science
Proposes a framework inspired by 'set and setting' to ethically harness placebo effects in psychedelic therapy, arguing that drug and nondrug factors have an interactive, synergistic relationship rather than being separable.
The difference between ‘placebo group’ and ‘placebo control’: a case study in psychedelic microdosing
Uses computational modeling to show how weak blinding plus positive expectancy creates 'activated expectancy bias' (AEB) that can inflate treatment effects; re-analysis of a microdosing trial suggests observed placebo-microdose differences are at risk of being false positives, and introduces the Correct Guess Rate Curve (CGRC) to correct for this.
Managing expectations with psychedelic microdosing
Highlights the role of expectations and confirmation bias in microdosing, noting that claimed therapeutic effects have little supporting empirical research, and calls for future studies to address this concern.