Self-blinding citizen science to explore psychedelic microdosing

This self-blinding experiment (n=191) finds that the placebo and microdosing groups both experienced similar improvements in self-rated psychological well-being and cognitive function (e.g. mood, energy, creativity) after four weeks. This study provides more evidence that microdosing benefits can be attributed to expectancy (placebo) effects.


Microdosing is the practice of regularly using low doses of psychedelic drugs. Anecdotal reports suggest that microdosing enhances well-being and cognition; however, such accounts are potentially biased by the placebo effect. This study used a ‘self-blinding’ citizen science initiative, where participants were given online instructions on how to incorporate placebo control into their microdosing routine without clinical supervision. The study was completed by 191 participants, making it the largest placebo-controlled trial on psychedelics to-date. All psychological outcomes improved significantly from baseline to after the 4 weeks long dose period for the microdose group; however, the placebo group also improved and no significant between-groups differences were observed. Acute (emotional state, drug intensity, mood, energy, and creativity) and post-acute (anxiety) scales showed small, but significant microdose vs. placebo differences; however, these results can be explained by participants breaking blind. The findings suggest that anecdotal benefits of microdosing can be explained by the placebo effect.

Authors: Balázs Szigeti, Laura Kartner, Allan Blemings, Fernando Rosas, Amanda Feilding, David J. Nutt, Robin L. Carhart-Harris & David Erritzoe


This paper was accompanied with a commentary/insight by Cameron (2021).

A related study on the effects of microdosing (without blinding) that showed that the expected effects were different from the actual effects was recently done by Polito and Stevenson (2019).

This study was also reported in New Atlas, Medical News, and The Guardian.

This study was unique in that it combined microdosing in a naturalistic (at home/normal life) environment with the blinding that usually happens in a clinical trial. This bridges the gap between large naturalistic studies (without blinding) and clinical trails that usually have a very small number of participants.

The participants (n=191) acquired their own microdoses (e.g. psilocybin-containing mushrooms 24% or LSD 61%) and with a clever QR system, they were randomized to three different conditions (placebo, half-half, microdosing). The microdosing part of the study lasted for four weeks. Each week the participants took four pills (placebo or up to two microdoses).

The participants were mostly male (70%), middle-aged (33.5), and healthy (33% indicated prior psychiatric diagnosis such as depression, 7% current (lower that the average population))

The following outcomes were measured at baseline, week five (after microdosing or placebo), and a follow-up at week nine:

  • Psychological well-being (RPWB)
  • Mindfulness (CAMS)
  • Life satisfaction (SWL)
  • Paranoid thoughts (GPTS)
  • Personality + intellect trait (B5)
  • Cognitive performance (CPS)

Every Sunday the following measures were taken:

  • Well-being (WEMWB)
  • Depressive symptomatology (QIDS)
  • Anxiety (STAIT)
  • Social connectedness (SCS)

And during the dosing period:

  • Positive and negative affect (PANAS)
  • Visual Analogue Scale (VAS, drug intensity, mood, energy, creativity, focus, and temper)
  • Cognitive performance (CPS)

The self-blinding was done in the following fashion:

“In summary, the two key elements of self-blinding are to hide the active components inside opaque capsules while preparing identical looking placebos (1) and to position non human-readable QR codes along the capsules prior to randomization (2). With the QR codes in place, it is possible for the experimenter to recover knowledge of capsule types after randomization without revealing that information to participants.”

The blinding was broken (breaking blind) at a higher rate than chance. This means that participants were often (72% vs 63% if random guesses) able to guess if they had taken a microdose. The higher the microdose, the more change the blind was broken (usually at 12 µg of LSD or higher). This information was also used in the analysis of the data.

The authors discuss what this study means in the discussion:

owever, when looking at the between-group comparisons of the same outcomes, no significant differences were found between the placebo and microdose groups. On the cognitive tests, which are less subjective than the self-reported psychological outcomes, the microdose group did not even improve from baseline to week 5 and the between-groups comparisons were not significant either. Thus, our study validates the positive anecdotal reports about the psychological benefits of microdosing (significant improvements from baseline in a broad range of psychological measures); however, our results also suggest that these improvements are not due to the pharmacological action of microdosing, but are rather explained by the placebo effect (lack of significant between-groups differences).

The study was limited in that it couldn’t verify the drugs being used (and how the microdosages were prepared). The authors also couldn’t follow the participants (see if they took the pills).

The current study was in a healthy population, and the authors note that a clinical population may possibly benefit from microdosing. But, when doing a further analysis of the current data, they didn’t find significant differences between those scoring highest and lowest on well-being scores.



This study used a self-blinding citizen science initiative to study the effects of microdosing on well-being and cognition. The results suggest that anecdotal benefits of microdosing can be explained by the placebo effect.

eLife digest

Psychedelic psychotherapy, therapy enhanced with psychedelic drugs such as LSD or psilocybin, has been suggested to improve psychological well-being. Recently, ‘microdosing’ has gained popularity as a way of administering psychedelics.

Modern medicine involves comparing two groups of patients: one that takes a drug and another that takes inactive pills, known as placebos. This ensures that observed effects are due to the drug itself.

To overcome the problem of expensive and difficult placebo-controlled studies on psychedelics, Szigeti et al. developed a new citizen-science approach, where microdosers implemented their own placebo control based on online instructions.

The trial examined whether psychedelic microdosing can improve cognitive function and psychological well-being. It found that both microdosing and placebo improved psychological measures, suggesting that the observed benefits are not caused by the microdose but rather by psychological expectations.

Psychedelic drugs are being used to treat depression, anxiety, addiction, and obsessive-compulsive behaviors.

Recently, ‘microdosing’ has emerged as an alternative paradigm of psychedelic use. It can be broadly defined as the frequent use of low doses of psychedelics.

Recent uncontrolled, observational studies have provided some empirical support for the idea that microdosing may improve well-being, creativity, and cognition. However, these studies are vulnerable to experimental biases and placebo effects.

A few recent double-blind, controlled studies have been conducted on microdosing. They found large variability in LSD blood concentration after microdosing, along with increased BDNF blood levels, but no robust evidence to support the positive anecdotal claims about microdosing.

Two issues need to be considered when assessing the scientific credibility of microdosing: the lack of placebo control in uncontrolled studies and the small sample size in controlled studies.

In the present study, we used a self-blinding setup procedure to allow participants to microdose on their own initiative and to implement placebo control and randomization without clinical supervision. We hypothesized that microdosing produces superior outcomes compared to placebo on psychological state and cognitive function.

Request a detailed protocol

This study had a naturalistic design involving elements of experimental control, prospective data collection and online citizen-science. It was 10 weeks long and had an optional follow-up at week 9.

Participants were given a placebo for 2 weeks and a microdose for 2 weeks.

Request a detailed protocol

In the accumulative phase, the following scales were assessed: Ryff’s psychological well-being, cognitive and affective mindfulness scale, satisfaction with life scale, green paranoid thought scales, big five personality traits, and intellect trait.

Participants were tested on six tasks to quantify cognitive performance: spatial span, paired associates, rotations, odd one out, spatial planning, and feature match. The cognitive performance score was calculated as the average z-score across the six tasks after removing learning effects.

Request a detailed protocol

Self-blinding was performed by packaging microdoses inside non-transparent capsules and placing them inside envelopes together with a QR code. Four envelopes were selected at random to correspond to the four weeks of the dose period.

The dose period started with an envelope being opened per week, and a QR code being scanned. This allowed us to deduce which type of capsule was taken when.

Self-blinding involves hiding the active components inside opaque capsules while preparing identical looking placebos, and placing non-human-readable QR codes along the capsules prior to randomization.

Participants were allowed to use any psychedelic substance to microdose with. The exact microdose dose was not defined, rather participants were instructed to use a microdose dose that they would use outside the study.

Request a detailed protocol

Participants were recruited through advertisement on relevant online and offline forums, and provided their email address and planned start date. They were required to have given informed consent and follow the study protocol.

Request a detailed protocol

All questionnaires were implemented online using the SurveyGizmo platform and Cambridge Brain Sciences. Individuals were sent links to each test via a dedicated email.

Request a detailed protocol

Participants were asked to guess the type of capsule they had taken that day. They were also asked to guess whether the current week was a microdose or a placebo week.

Request a detailed protocol

Accumulative outcomes were analyzed with mixed-effect repeated measurement models, and between-group comparisons were made at week 5 and week 9. Additionally, within-group comparisons were made at week 5 and week 9 between PL and HH and PL and MD.

To analyze acute and post-acute outcomes, mixed linear models were constructed with score as dependent variable, subject ID as a random-effect, and condition as fixed-effect. Planned comparisons were made between scores obtained under placebo/microdose week and without placebo/microdose week.

To better understand how guess influenced scores, models were constructed with the addition of guess and guess*condition factors. Finally, post-hoc comparisons were made between PL and MD conditions in 4 strata (condition/guess).

Request a detailed protocol

The study engaged people who planned to microdose on their own initiative, and who consented to incorporate placebo control into their self-experimentation. Email addresses were the only personally identifiable data collected.

Demographics, randomization, and completion rate

A total of 1630 participants signed-up, 240 started the study, and 191 completed it. There were no statistically significant differences between the three groups in any demographic, recreational drug use or baseline measures.

The sample consisted of educated, middle-age, healthy males from western countries, with a positive attitude toward psychedelic drugs, in particular toward medical use. Only 7% of participants had a current mental diagnosis.


Most participants used LSD or analogues, followed by psilocybin containing mushrooms, and three individuals used other psychedelics.

Accumulative outcomes

Accumulative outcomes were collected at baseline, week 5 and week 9, and compared within and between groups at the week 5 and week 9 timepoints.

All self-reported psychological outcomes improved significantly in the MD group from baseline to week 5, including well-being, mindfulness, life satisfaction, and paranoia. Personality structure also improved in the MD group, including reduced neuroticism trait score and increased openness. In the MD group, rotations and odd one out scores increased, while spatial span and paired associates scores decreased. These increases were sustained at the follow-up.

Planned comparisons revealed no significant between-group differences at the week 5 or week 9 follow-up timepoints, except that the paired associates scores decreased in the HH group.

Accumulative outcomes adjusted for number of microdose guesses

We performed a post-hoc adjustment for the number of microdose capsules guessed to further examine the role of placebo-like expectation effects in the accumulative outcomes. The adjustment decreased the already small between-group differences on self-reported scales, while it did not affect cognitive outcomes.

Acute and post-acute outcomes

Results are presented without considering the guess component, which is discussed in the next section. The average sample size was 857 for psychological measures and 684 for cognitive performance.

Among acute measures, condition (PL vs. MD) was significant for acute emotional state (PANAS), acute drug intensity (12.5, 3.0), acute mood (4.6, 2.9), acute energy (5.3, 2.7), and acute creativity (4.7, 2.6) VASs.

Association between guess and acute/post-acute outcomes

The acute and post-acute results were re-analyzed with guess included. The guess*condition interaction term was non-significant for all scales, except for drug intensity.

To better understand the role of guess in self-reported outcomes, two binary variables were used: condition (PL/MD) and guess (PL/MD). When condition was fixed, significant differences were found in 21 of 22 comparisons, all favoring MD guess.

Blinding integrity

Participants had a higher break blind rate (0.72), higher specificity (0.82) and higher sensitivity (0.46) than a random guesser (0.62) with knowledge of the ratio of PL/MD capsules (3/1). This suggests that the higher break blind rate is mostly due to superior ability to identify microdoses.

Break blind rate was positively associated with reported microdose dose. The estimated detection threshold was 12 g.


We used a novel self-blinding methodology to investigate the acute, post-acute, and long-term effects of repeated psychedelic microdosing.

When looking at changes over time from baseline to week 5, the microdose group showed significant improvements in well-being, mindfulness, life satisfaction, and paranoia, but no significant differences were found between the placebo and microdose groups on the cognitive tests.

The addition of the guess variable to the models reduced the significance of the effects on post-acute anxiety, acute emotional state, mood, energy, creativity, and drug intensity.

When the guess was fixed, there were no significant differences between placebo and microdose conditions, except for drug intensity (MD>PL). However, when condition was fixed, there were significant differences between placebo and microdose conditions in 21 out of 22 comparisons, always favoring the microdose guess.

In this study, participants guessed their capsules correctly in 72% of the cases, which is higher than random, but not as high as reported in antidepressant studies. This may be due to non-specific treatment factors such as expectation of a benefit and investigator alliance. The acute and post-acute results observed in this study could be explained by the difference in expectations between participants when a microdose is perceived versus the absence of expected benefits when placebo is perceived.

Blind breaking was induced by increased drug intensity, which was mostly manifested as body and perceptual sensations. This finding suggests that blind breaking induced clinically irrelevant side effects, rather than deduced from improvements of outcome variables.

The current study used a 3-arm design to protect blinding integrity, and participants in the placebo group were in disbelief after opening their unused envelopes containing unused capsules after the conclusion of the study.

I counted the number of cut blotters I had in the left overs, and they are 8…so you must be right. I was indeed taking placebos throughout the trial.


This study is part-controlled, part-observational and yields data superior to conventional observational data, but inferior to controlled clinical trial data.

The present study is limited by the lack of verification of the nature, purity, and dosage of the psychedelic substance used for microdosing. However, the positive relationship between dose and blind breaking (Figure 4) and the consistent threshold dose for psychoactivity provide some reassurance.

We could not confirm whether participants followed the self-blinding procedure accurately, and we had no way of confirming whether the capsules were taken as instructed during the dose period.

We cannot determine a strict causal relationship between guess and outcome, because guess was recorded after completion of assessments, but most participants reported to break blind due to body and perceptual sensations, rather than improved outcomes.

We cannot rule out the possibility that a clinical sample would have yielded more promising results, but the present healthy sample has less scope for potential improvements.

In this opportunistic naturalistic study, participants with the lowest 25% baseline well-being scores and those with the highest 25% baseline neuroticism scores received the same amount of microdosing as the complete sample, and there were no significant differences between conditions for any of the accumulative outcomes.

Although the MD group improved more than the PL group on all scales (from baseline to week 5), the difference was not statistically significant. The sample size required to detect a true between-group difference would be too small.

Future directions

The successful execution of this citizen science initiative may inspire similar initiatives throughout the world in a broad range of scientific and medical contexts. The initiative is low cost and provides incorporation of randomization and placebo control.

The current study suggests that selecting dosage for future microdosing trials is fraught with difficulties, and that careful assessment of blinding could alleviate some of these concerns.

The present study has implications for full/’macrodose’ psychedelic studies, where blinding is impossible due to the intense nature of the experience. It raises interesting philosophical and ethical questions, and might also suggest that belief is an active component of the psychedelic treatment model.

Here we created a novel methodology for conducting a placebo-controlled study on psychedelics. The results suggest that the benefits of microdosing are not due to the pharmacological action of the drugs.

Participants rated items on a 5-point Likert scale to determine their score on the 5-factor model of personality.

Cognitive performance score (CPS)

Cognitive performance was measured in six tasks: spatial span, paired associates, rotations, odd one out, spatial planning, and feature match. A single cognitive performance score (CPS) was calculated to quantify overall cognitive performance.

The CPS score is the difference between the placebo group and the average of the six subtasks.

Daily effects of microdosing VASs (DEMS)

Please rate your mood, your energy, your temper, your ability to keep focused, and your creativity for today.

Green paranoid thought scales (GPTS)

The 16-item ‘social reference’ subscale was used to assess paranoia. Higher scores indicate higher levels of paranoia, and lower scores indicate improvements.

Positive and negative affection scale (PANAS)

PANAS is a 20-item scale that consists of words that describe different feelings and emotions. It has two subscales, positive and negative.

Quick inventory of depressive symptomatology (QIDS)

The 16-item self-report version of the scale was used, and lower scores indicate fewer depressive symptoms.

Ryff’s psychological well-being (RPWB)

A 42-item instrument was used to quantify well-being as a single outcome. A seven-step rating was used by accident, and scores were rescaled by multiplying them with 6/7 and rounding it to the closest digit.

Short suggestibility scale (SSS)

The SSS is a 21-item, unidimensional scale that quantifies an individual tendency to accept messages.

Spielberger’s state-trait anxiety inventory (STAIT)

A 20-item scale was used to assess how often participants felt certain feelings or mental states.

Warwick–Edinburgh mental well-being (WEMWB)

A 14-item unidimensional scale covering both the feeling and functional aspects of mental well-being was used in analysis.

Additional information on the self-blinding setup and data collection

For the MD group, all 4 weeks are microdose weeks, but the order of the MD weeks could be in any order, and for the HH group, any two of the MD weeks could be used.

If psilocybin containing mushrooms or liquid was used, placebo capsules had to be filled with equal weight of non-psychoactive mushrooms, chaga (Inonotus obliquus) was recommended.

Participants received an automated report after the completion of the long-term follow-up timepoint. This report did not affect any outcome measures.

Twelve participants chose to construct their own dosing schedule with minimal restrictions, and seven of them maintained two microdoses/week (on MD weeks) and only moved the days of microdoses. Data from five individuals was excluded to simplify the analysis.

Three participants had invalid dose schedules, likely due to errors during the setup. Their data were discarded from the between-group analysis, but their acute and post-acute measures were used.

In general, participants had positive attitudes toward psychedelics and consisted of knowledgeable users. The mean of three VAS items was calculated to measure participants’ expectations about microdosing, and the sample had highly positive expectations about microdosing.

Blind breaking cues

After the data collection was closed, a short survey was conducted among participants. The results are summarized in Table 1.

Participants described muscle tingling/tension, increased color saturation/visual warping, and stomach stress/strain in the non-mandatory text box below the question. Only three text responses included effects related to the outcome measures.

Comparison to random guesser

The random guesser only knew the ratio of PL/MD capsules (3/1) in the envelopes, and this ratio remained constant through the experiment.

To calculate the threshold dose, 200 simulations were conducted with random guessing and the resulting distribution of random guess accuracies was tested against the participant’s guess accuracy. The threshold dose was found to be 12 g.

Decision letter after peer review:

Thank you for submitting your article to eLife. Three peer reviewers have reviewed it and the evaluation has been overseen by Alexander Shackman and Chris Baker.

The editors have judged that your manuscript is of interest, but additional experiments are required before it is published. We will give authors as much time as they need to submit revised manuscripts, and we will post the manuscript to bioRxiv if you choose.

A complex, ingeniously designed online study was undertaken to examine the acute, post-acute, and enduring effects of psychedelic microdosing.

At first, I had my doubts about self-blinding, but became very enthusiastic about this method as it is well thought-out and a good addition to the literature.

o More clarity is needed about how participants were blinded and guided to know what a microdose was.

We have added a new subsection to the Materials and methods that provides a high-level overview of the self-blinding setup procedure.

We have added a new subsection to the Materials and methods, detailing the directions given to participants. The new section reads: “Participants were allowed to use any psychedelic substance to microdose with.”

In the study, data when participants breaking blind are included in all analysis, in agreement with conventional psychiatric trials that also include such data.

We asked participants to guess their treatment daily, so we can only talk about the % of days when participants break blind, rather than the % of participants.

The data together with the documentation is publicly available in the data/ folder of the repository we shared. The Transparent reporting document has been updated to point to the data folder.

Study details

Compounds studied
LSD Psilocybin

Topics studied
Anxiety Healthy Subjects Creativity

Study characteristics
Placebo-Controlled Single-Blind


Linked Research Papers

Notable research papers that build on or are influenced by this paper

The difference between ‘placebo group’ and ‘placebo control’: a case study in psychedelic microdosing
This computational analysis (2023) employs computational modelling to illustrate how weak blinding and positive treatment expectancy can cause an uneven distribution of expectancy effects, termed 'activated expectancy bias' (AEB). The study introduces the Correct Guess Rate Curve (CGRC), a tool to estimate the results of a perfectly blinded trial using data from an imperfect one, and re-analyzed the ‘self-blinding psychedelic microdose trial’ dataset (n=191) to demonstrate that placebo-microdose differences may be susceptible to AEB, thereby suggesting microdosing can be viewed as an active placebo.

Evidence for tolerance in psychedelic microdosing from the self-blinding microdose trial
This pre-print reanalysed data from a placebo-controlled citizen science microdosing study (n=240) to investigate whether tolerance develops during microdosing. It was conceptualized that if tolerance develops, the probability of correctly guessing active microdoses should decrease with more microdoses. Linear regression models showed that correct microdose guess probability decreased with the number of microdoses taken (p=0.09) indicating that tolerance developed.

PDF of Self-blinding citizen science to explore psychedelic microdosing