Microdosing — what the peer-reviewed research actually shows

Microdosing refers to taking sub-perceptual or barely-perceptual doses of psychoactive compounds, typically psilocybin or LSD, on a regular schedule (commonly 1-on / 2-off, after the Fadiman protocol). "Sub-perceptual" means below the dose at which a typical user would notice acute effects.

Practical dose ranges in the literature:

• Psilocybin: 0.1-0.5 g dried mushroom (~1-5 mg psilocybin)
• LSD: 5-20 µg (about 1/20th to 1/5th of a typical recreational dose)

The premise: chronic low-dose exposure produces gradual changes in mood, creativity, and cognitive flexibility without the acute experience of a full-dose session.

We are not in a position to recommend microdosing. Both psilocybin and LSD are Schedule I federally. This article describes what the published research has shown and helps you evaluate the literature critically.

Microdosing presents three serious methodological challenges that any reader of the literature needs to understand:

1. Self-selection bias. People who choose to microdose are not randomly sampled from the population. They tend to be psychologically open, somewhat psychedelic-experienced, and motivated by specific outcomes (depression, creativity, focus). Naturalistic studies of microdosers (e.g., Polito 2019, Anderson 2019) report meaningful improvements — but those improvements may reflect who chose to microdose, not what microdosing did.

2. Placebo effects are large in psychiatric outcomes. Antidepressant trials routinely show 30-40% placebo response. Self-perceived improvements in mood and cognition are particularly susceptible to expectation effects.

3. Self-blinding is difficult. A user generally knows what dose they took. Even if "sub-perceptual," many users report being able to feel a microdose enough to break blinding.

The studies that have addressed these challenges with proper controls have produced more modest results than naturalistic data alone suggested.

Szigeti et al. (2021), published in eLife, is the largest pre-registered self-blinding study to date. 191 participants who were already planning to microdose were randomized to dummy, low-dose, or alternating-dose protocols using opaque capsules they prepared themselves. The mood and cognitive outcomes were measured at 4 weeks.

Results: participants who believed they had microdosed (regardless of what was actually in the capsule) reported substantial improvements. But there was no significant difference between actual microdose and placebo arms.

This suggests the substantial effects reported in naturalistic studies may be largely placebo-mediated.

Marschall et al. (2022) — a similar pre-registered design — replicated the placebo finding for some outcomes but found small drug-specific effects on certain creativity measures.

The honest summary: naturalistic studies overstate effects substantially. Pre-registered self-blinding studies suggest the drug-specific effect is either small or restricted to specific subdomains.

Polito & Stevenson (2019, PLoS ONE) followed 98 microdosers over 6 weeks using ecological momentary assessment. Results: improvements in some self-rated outcomes (creativity, focus); decreases in others (neuroticism). Statistically significant changes for several outcomes; effect sizes were small to moderate.

Anderson et al. (2019) compared 909 self-identified microdosers to 233 controls cross-sectionally. Microdosers scored higher on creativity, openness, wisdom, and lower on neuroticism. The cross-sectional design can't establish causation — and the self-selection issue is acute.

The Imperial College team (Kuypers et al. 2019) ran a controlled lab study with single-dose 0.7 µg/kg LSD. Found improvements in convergent thinking; no significant effect on divergent thinking. This is a single-acute-dose study, not chronic microdosing — but it gives some signal that small doses may have measurable effects on specific cognitive subdomains.

Putting the above together, the cautious read of the literature in 2026 is:

Plausible but not established:

• Modest mood improvements over weeks-to-months of microdosing
• Modest improvements on specific creativity or cognitive flexibility measures
• Small but real drug-specific effects beyond placebo

Not supported:

• Dramatic transformations
• Specific therapeutic claims for depression, ADHD, etc. (these may be true but the evidence isn't there yet)
• Long-term safety claims either way (chronic-dosing pharmacology is genuinely unknown)

Genuine concerns:

• Cardiac risk: psilocybin and LSD both have 5-HT2B agonist activity, which is the receptor associated with valvular heart disease in long-term high-dose use of fenfluramine and cabergoline. Whether chronic microdosing produces clinically meaningful cardiac effects is unknown but theoretically plausible.
• Tolerance: psilocybin tolerance develops within hours and persists 5-7 days; this is part of why protocols include rest days.
• Drug interactions: SSRIs may attenuate psychedelic effects; MAOIs may dangerously potentiate them.

If you are evaluating microdosing claims:

• Be skeptical of testimonials — they're heavily affected by expectation.
• Look for pre-registered self-blinding designs in the literature; weight those most heavily.
• Recognize that the gap between naturalistic effect sizes and controlled effect sizes is large — typically 3-5x.
• Take the cardiac risk question seriously; long-term chronic dosing has not been studied.
• Remember the legal context: psilocybin and LSD remain Schedule I federally.

The science is not settled. The honest position is "there's a small drug-specific effect that may or may not be clinically meaningful, plus a large placebo effect that is real and useful but not what people typically attribute it to." That's a much less exciting headline than "microdosing transforms cognition" — it's also closer to what the controlled literature shows.