Creativity

Last reviewed 5 October 2026

A note on the compounds. Some of the controlled work described here used LSD, which is semi-synthetic, or a laboratory formulation modelled on a plant preparation (a DMT and harmine combination inspired by ayahuasca) rather than the plant preparation itself. Where a study used one of these, the text says so. On this page, "full-dose" means an amount large enough to produce the complete psychedelic experience, and "microdosing" means repeated, much smaller amounts that are not meant to produce it. No dose figures appear anywhere below.

What it is

In the laboratory, creativity is usually split into two kinds of thinking. Divergent thinking is the ability to produce many different ideas from one starting point, as in the Alternate Uses Task, where a person lists unusual uses for a brick. Convergent thinking is the ability to narrow a field of possibilities to the one that works. The Remote Associates Task tests it by giving three words and asking for the single word that links them. The Picture Concept Task, which asks for the one idea shared by a grid of pictures, can be scored for both modes. A fourth method, the Consensual Assessment Technique, asks experts to rate something a person has made. It comes closest to real creative output and has rarely been used in this field.

Three more distinctions run through the research. The first is timing. What happens during a session can differ sharply from what shows up days or weeks later, and the studies below keep these apart. The second is deliberate against spontaneous creativity. Deliberate creativity is what a timed task measures. Spontaneous creativity is the sudden sense of an insight, and studies usually capture it only through a questionnaire [2]. The third is reported against measured creativity. People can feel more creative without scoring any differently on a task, and the gap between the two is one of the main findings of this literature.

Meta-analyses of creative achievement against intelligence test scores and against divergent thinking scores, covering 17 and 27 studies (5,544 and 47,197 participants), found relationships that were statistically significant but marginal in size (r = 0.216 for divergent thinking, 0.167 for IQ) [1]. In one trial, vocabulary score also significantly affected results on the verbal tasks [3]. A change on a ten-minute task is therefore a small and indirect signal about a person's creative life.

Many people assume psychedelics make them more creative, and much popular interest rests on that belief. Creativity also has strong non-drug comparisons, so an honest account has to show how the psychedelic evidence compares with walking, training and sleep.

What is known

The short version. In healthy adults, controlled trials of full-dose psilocybin, LSD and an ayahuasca-inspired compound show a fall in convergent thinking during the session, with mixed results for divergent thinking. Controlled trials of microdosing mostly find no change in creativity task scores compared with placebo. Studies without a placebo, and surveys, report friendlier results. One psilocybin trial found a borderline rise in one divergent measure a week after the session, and we did not find a second controlled trial reporting it.

Full-dose sessions in controlled trials

A randomized, double-blind crossover study of 24 healthy volunteers gave each person LSD or placebo and applied a battery of five creativity tasks near the peak of the drug's effect. The authors describe three patterns: a "pattern break" (higher novelty, surprise, originality and semantic distance between ideas), a loss of organization (lower utility and lower convergent thinking, and marginally lower elaboration), and "meaning" (more symbolic thinking and ambiguity) [4]. LSD is semi-synthetic.

The psilocybin trial is the most informative single study. It randomized 60 healthy adults to psilocybin or placebo in parallel groups, double-blind, and tested people before the session, during the peak and seven days later. All had taken a psychedelic before, though not in the preceding three months. During the session, convergent thinking fell (Cohen's d = 0.85), and so did the number of ideas on both divergent tasks (d = 0.84 and 0.80). Originality fell on the picture task (d = 0.65), but the ratio of original ideas to all ideas did not differ from placebo, so the fall mostly reflects fewer ideas rather than poorer ones. Yet participants rated their own state as more insightful, on three self-report items completed about six hours after dosing, which the authors read as a gap between feeling and performing or as a split between spontaneous creativity, which rose, and deliberate creativity, which fell [2]. Seven days later, the psilocybin group listed more ideas they had never thought of before (d = 0.52), but total ideas and originality did not differ from placebo. Convergent thinking was still lower than in the placebo group (d = 0.60), although the placebo group improved with practice during the session while the psilocybin group did not improve until follow-up, so part of that gap may be a blocked practice effect. Two cautions apply. The novelty result was borderline (p = 0.05) and was the only sign of a rise among the eight behavioural measures at follow-up, and the paper applies no correction for multiple comparisons [2].

The DMT and harmine study used a double-blind, placebo-controlled, within-subject design with 30 healthy men, who each attended three sessions: an ayahuasca-inspired DMT and harmine formulation, harmine alone, and placebo. The combination impaired convergent thinking, especially in people with higher baseline reasoning. Divergent thinking showed no overall effect, with trend-level reductions in fluency and elaboration. Subjective insightfulness predicted divergent outcomes but not convergent ones [5]. This was a laboratory formulation, tested in young men only.

Put side by side, three trials of three compounds agree on one thing: convergent thinking drops while the effect lasts. Divergent thinking does not follow a single pattern: it rose on novelty measures with LSD, fell in idea count with psilocybin and stayed flat with DMT and harmine.

What persists after the session

Evidence on later effects is thinner than evidence on acute ones. Besides the seven-day result above, three studies followed people after real-world sessions with no placebo group. At a psilocybin retreat, 55 people were tested before, 50 the morning after and 22 at seven days. Divergent thinking (fluency and originality) was higher the morning after, and the gains were gone by day seven. Convergent thinking did not change the morning after and was higher only at seven days, while life satisfaction was higher at both points [6]. At ayahuasca retreats in the Netherlands (43 people at baseline, 20 the morning after, 19 at one week), only the ratio of original ideas to all ideas fell, at both follow-ups. Raw originality and idea counts did not change. Convergent thinking was unchanged the morning after and higher at seven days (d = 0.83), the same late rise seen at the psilocybin retreat, which the authors say practice may explain [7]. A field study of 26 experienced ayahuasca users at two workshops, tested 1.5 to 2 hours after the first dose, found convergent thinking down on the picture task and, of the divergent measures, only the original-to-total ratio up. Raw originality did not change, raw idea counts moved in opposite directions in the two groups, and the groups differed in age [8].

These studies disagree with each other, and none had a control group. Two found convergent thinking higher at seven days, which sits against the controlled psilocybin trial, where convergent scores were lower than placebo at that point. The authors of the controlled trial say their result calls the naturalistic findings into question, and none of the uncontrolled studies could rule out a practice effect [2]. Retreat attendees chose to be there, and only 40 percent of the psilocybin retreat group (22 of 55) and 44 percent of the ayahuasca group (19 of 43) remained at follow-up.

Reviews describe the same uncertainty. A systematic review of psilocybin and cognition (20 articles, 2,959 participants, 85 percent of the studies in healthy volunteers) concluded that cognitive flexibility and creative cognition seemed to decline at first and might improve with time. The hopeful part rests on a self-reported flexibility rating in one online survey, which supplies most of the 2,959 participants, and on an open-label sample of people with depression. In the one controlled creativity trial it covered, convergent thinking was still lower at day seven. The review covers cognition in general, not creativity alone [9] [2]. A scoping review of 42 studies found that full doses tended to impair creativity during the session, while the few studies with later measures (one to 85 days afterward) were mostly null with some positive results. It also reported that microdoses tended toward creative enhancement in the acute phase, which differs from the controlled microdosing data below, but that is a count of 18 findings, 12 neutral and 6 positive with none negative, mostly from uncontrolled studies. The authors call their findings preliminary, warn of possible publication bias, and searched to early 2023 [10]. An integrative review of 11 studies suggested that convergent thinking may rise in the post-acute period, but rests on few studies, small samples and almost no randomized trials [11]. A 2026 systematic review of 67 studies (47 in people, 20 in animals, covering compounds such as ketamine and MDMA) found psilocybin the most consistent for later gains in cognitive flexibility and attention, while LSD and DMT showed no consistent change. The gains were seen more in psychiatric populations than in healthy people, which the authors suggest may be remediation of an existing deficit rather than enhancement. That review measured executive function, attention and decision-making, not creativity [12]. A review of 34 experimental studies with follow-up of at least two weeks documented lasting changes in personality and well-being, but its abstract names no creativity outcome [24].

Our reading is that a later benefit is possible but weakly supported. It rests on one controlled trial, one borderline measure and no replication, and nothing has shown it to last.

Microdosing in controlled trials

The largest synthesis is a 2026 systematic review and meta-analysis of 24 studies and 3,681 healthy participants. Only six studies entered any meta-analysis, and the randomized comparisons pooled in the abstract concern depression, anxiety and stress in 117 people, so the statement about creativity appears to be a narrative one. It reads that randomized trials showed "minimal impact on cognition, creativity, or sustained mood." Observational studies reported mood and personality improvements that the authors attribute partly to expectation and lifestyle [13].

The most thorough single creativity trial randomized 80 healthy men to six weeks of repeated very low LSD doses or placebo. Creativity was tested at baseline, 240 minutes after the first dose, and at a final visit two days after the last dose. The first test fell past the drug's peak, though the effect was still about 82 percent of its maximum. The battery had a divergent task, a convergent task, a collage design that 16 art teachers rated, and a problem-solving questionnaire built for the study, with no test-retest reliability reported. No measure showed a drug-by-visit effect. The authors applied no correction for multiple comparisons and call the analyses exploratory. Seventy-five of the 80 finished all doses.

Three double-blind, placebo-controlled trials of psilocybin truffles from one university group (one preregistered) were pooled for 171 people. Active microdosing raised the ratio of original responses to total responses, which the authors read as higher quality of divergent thinking. These effects held after controlling for participants' guesses about their condition. No other divergent score and no convergent score changed [14]. We could not read the published paper. An earlier preprint of the report (not peer reviewed) describes the pooled analysis as exploratory, run after the first two trials disagreed. In the preregistered third trial the main interaction was only borderline (p = 0.053) [33]. A companion report on two of the trials found no significant effects on cognitive control, memory, social cognition or well-being after correcting for multiple comparisons, and participants stayed blinded [15]. The creativity result is therefore one narrow signal from one group, and we did not find an independent replication.

Other blinded studies point the same way. A double-blind trial of 34 people who were starting to microdose with psilocybin mushrooms found stronger acute effects only in those who correctly guessed their condition. The condition became apparent to participants in 49 of 64 weeks (75 percent). The trial found no support for better creativity, well-being or cognition, and a few small changes toward impairment [16]. A self-blinding citizen-science study of 191 volunteers, in which participants ran their own placebo control and the substances were not verified, found no difference between the active and placebo groups on any of the longer-term measures. Both groups improved on some of them, and the active group on more. Creativity was one self-rated item, not a task. On dosing days it was 4.7 points higher with the active substance, and the gap fell to a non-significant 1.8 once participants' guesses about their condition were taken into account [17]. In 20 healthy volunteers, single very low LSD doses left cognition unaffected, and the only creativity measure, the convergent Remote Associates Task, showed just a marginal rise in items attempted, so the study says nothing about divergent thinking [18].

Studies without a placebo, and surveys

The study that started much of the public interest tested 38 volunteers from about 80 attendees at a microdosing event. People were tested before and roughly 90 minutes after taking psilocybin truffles. Analyses covered 27 people on the convergent task and 33 on the divergent task. Convergent scores rose (d = 0.49), and fluency, flexibility and originality scores rose too, while a fluid intelligence test did not change. There was no control group and no blinding, the sample was self-selected, and agreement between raters was only fair to moderate (kappa .23 to .55). The authors' argument against a practice effect rests on earlier literature, and the controlled psilocybin trial later found one on the same picture task [2]. The authors asked for placebo-controlled replication [19]. The later blinded trials from the same group did not reproduce the convergent improvement [14] [15].

Other designs lean the same direction. In a survey of 68 people, the 11 who reported psychedelic use alongside a mystical experience solved a functional-fixedness puzzle faster (d = -0.87), though the mystical experience itself was unrelated to speed [20]. An online comparison of 187 psychedelic users with 139 non-users found higher originality and fluency in users, but the groups differed in age, sex and psychiatric diagnosis (37 percent of users against 25 percent), so the users were not a healthy-only group. More frequent use related to originality only as a trend, and real-life creative achievements and well-being did not differ [21]. A prospective survey of 1,435 people who microdose found higher self-rated creativity on dosing days than on other days (F = 25.99), and its authors urge caution because the design is observational [22]. Daily ratings from 98 microdosers over six weeks rose on dosing days with little carry-over, and all seven rated items rose together, which looks like a general lift rather than a creativity effect. A trait measure of creative personality did not change. A second study of 263 people found that creativity ranked first of 16 among expected benefits and ninth of 16 among reported ones [23]. A 2025 questionnaire validation in 1,323 recent psychedelic users includes an "inspiration and creativity" factor, but it records self-reported experience in people who chose to take a psychedelic, the comparison group was users of stimulants or opioids rather than non-users, and there was no measurement before use [32].

None of these designs can separate the drug from who chooses to take it, what they expect or the setting. The more favourable the result, the less likely the study was blinded, though not uniformly. The ayahuasca workshop study found convergent thinking fell and only an idea-quality ratio rose [8], and the ayahuasca retreat study found that same ratio fell [7].

Reported creativity and measured creativity

Ratings of insight rose while idea counts fell in the same people in the psilocybin trial [2]. In earlier reporting from the same LSD microdosing trial, self-rated creativity rose on dose days with no task change, and an exploratory analysis found that rise absent on days when participants were unsure which they had taken [3]. The authors of the LSD trial suggest that the feeling of creativity may track a general lift in mood and an inflated sense of meaning, and leave open that real effects exist but sit outside what laboratory tasks capture [3]. What the data support is that feeling more creative and producing more creative work are different outcomes, and in the trials above the two often moved apart.

What the creativity studies say about harm

All of these trials ran with screening and supervision, so they do not describe unsupervised use. Of the 42 studies in the psilocybin scoping review, 26 percent reported safety outcomes at all, and one of those reported serious adverse reactions, in people with treatment-resistant depression rather than healthy adults [10]. In two microdosing trials pooled for adverse events (109 people), risks were similar to placebo [13]. In the truffle trials, negative bodily feelings were more common in the active group [15]. One cognitive effect is itself a safety-relevant finding: convergent thinking, the ability to judge which idea works, was lower during a full-dose session in all three controlled trials [2] [4] [5]. This page does not assess other risks.

What is not known

Most of the questions that matter for this dimension are open.

No study we found tests whether one well-supported full-dose session changes real creative output, as opposed to task scores, weeks or months later in healthy adults. The longest controlled follow-up in this literature is seven days [2]. We also did not find a study that paired a session with deliberate creative practice. The authors of the psilocybin trial raised that idea for therapy and left it untested [2].

Measurement is the second gap. Most tasks were built for students and for detecting differences between groups, not for detecting improvement in people who are already doing well. In the DMT and harmine study the impairment was largest in people with higher baseline reasoning [5]. Scoring itself is shaky: raters agreed only modestly in an open-label study [19], and the placebo group in the controlled psilocybin trial improved with practice [2], so a change in a score can reflect practice or the rater as much as the drug. Product-based measures of real output are rare: the integrative review found only one instance among its studies [11], and the expert-rated collage in the LSD microdosing trial was the first in a microdosing study [3].

Third, the two positive controlled signals are fragile and need independent replication: the borderline rise in novel ideas a week after psilocybin [2] and the higher share of original responses after truffle microdosing, from an exploratory pooled analysis [14] [33]. Each comes from a single research group and a modest sample.

Fourth, full-dose and microdose results should not be pooled. They show different patterns, and treating them as one question blurs both.

Last come setting and expectation. The LSD microdosing authors suggest that laboratory conditions may dampen effects, and the favourable event study had a social setting and unblinded participants [3] [19]. We found no study that tested creativity in realistic settings with a placebo group. The existing trials also drew mainly young volunteers, and two enrolled only men [3] [5].

These are open questions Metanoia wants to help answer. None of them is an announced study. In its own words: Metanoia is a California 501(c)(3) public charity advancing the science of natural psychedelics for personal growth and human flourishing in healthy adults. Not treating illness. Expanding potential.

Compared with other paths

The evidence for other paths also mostly measures task scores soon after the activity, and much of it comes from students. Effect sizes from different designs are not directly comparable. With that said, the alternatives are better studied.

Creativity training. A 2023 meta-analysis of creativity enhancement in adults pooled 332 effect sizes from 84 studies. The overall effect was medium (Hedges' g = 0.53). Most outcomes were measured the same day, but 28 effect sizes taken more than 100 days later averaged g = 0.74, so training effects can last. The spread was wide, and the authors say a single overall average means little for that reason. Idea quality, rarely assessed, showed almost no gain. The search ran from January to March 2019, before the recent psychedelic trials [25].

Meditation. The same meta-analysis found meditation at g = 0.66 across 19 effect sizes [25]. A small controlled study of 19 experienced meditators found that an open-monitoring session raised divergent thinking compared with a focused-attention session and a baseline session, with no clear effect on convergent thinking. Mood rose after both kinds of meditation, so the authors cannot say whether mood drove the result, and they note possible demand effects [26]. The Mindfulness page covers that literature.

Walking and physical activity. A meta-analysis of 23 studies and 1,036 participants, mostly students, found a large effect of walking on divergent thinking (d = 0.93, 95% CI 0.44 to 1.42), and d = 0.82 (0.35 to 1.28) when only randomized studies were counted. The authors grade certainty for the 13 randomized studies as moderate, report high heterogeneity, and note that many positive studies came from one author. The search ran to January 2025. Convergent thinking showed no clear effect (d = 0.16, CI -0.31 to 0.63, very low certainty) [27]. Walks lasted 2 to 20 minutes in all trials but one, so this is a rough comparison with the acute psychedelic data. Physical activity as a broad category did less well in the training meta-analysis (g = 0.26, CI -0.14 to 0.65) [25], so the walking result should not be stretched to all exercise.

Time in nature. In one study, 56 adults on wilderness trips took a ten-item Remote Associates test, either before the trip (24 people) or on day four (32 different people). The later group solved about 6.1 items against 4.1 (d = 0.86) after adjusting for an age gap of about ten years, and whole hiking groups were assigned to one test or the other. The authors cannot tell whether nature or the absence of technology produced it [28]. It is a single study with a convergent task, and we did not look for replications.

Sleep. In a Nature study, more than twice as many people found a hidden rule in a task after a night of sleep as after the same time awake [29]. In a nap study, people who entered REM sleep improved on remote-association problems, but only on items primed earlier in the day. Groups were defined by what happened during the nap, and only 12 people had non-REM sleep [30]. In a similar hidden-rule task with a 20-minute rest, 20 of 24 people who reached the lightest sleep stage (83 percent) found the rule afterwards, against 15 of 49 who stayed awake (31 percent), in groups sorted by what happened during the rest and in young easy sleepers who had slept less than usual the night before [31]. These insight problems, tested in laboratory samples, show that sleep can change how people solve problems and do not show a general rise in creativity.

Drugs for cognitive enhancement. In the training meta-analysis, the drug category (alcohol, Adderall and modafinil, with no psychedelics among them) averaged g = 0.10 across 20 effect sizes, which was not significant [25].

Where psychedelics fit. On present evidence in healthy adults, psychedelics have not been shown to raise creativity in any lasting way. Walking and training courses have larger, better-replicated effects on task scores, with the limits noted above. No study has tested a psychedelic head to head against any of these. The psychedelic state does show a documented change in how people think and in their sense of insight, but whether that has practical value for creative work is an open question, and this page does not claim an advantage over any other path.

History and tradition

These are stories and early reports, described through peer-reviewed papers. They are not evidence that any substance raises creativity.

Reports that psychedelic experiences inspire creative work are old and widely repeated. One of the controlled trials above lists several famous stories, among them a celebrated discovery in molecular biology, the 1960s California computer industry, and works by well-known authors, and it calls them anecdotal [2]. This page does not repeat or verify any of them. Researchers began testing the idea in the 1950s, and that work ended when the drugs were banned in the United States in the late 1960s. The same authors say the findings were largely inconclusive by present standards [2].

The book on all twelve dimensions   Back to the twelve dimensions