The effects of psilocybin on time perception in humans: A comparative analysis of subjective and objective measures

Psilocybin has regained scientific and clinical interest, but its effects on temporal perception in humans remain relatively under-studied. Earlier psychedelic research had shown that psychedelics can alter the experience of time, yet most human work on psilocybin had relied on temporal reproduction or production tasks, leaving it unclear whether similar effects would appear in a temporal bisection task, which places different demands on memory, attention and motor responding. The authors also note that the role of the serotonergic system in timing is still uncertain, and that subjective reports of altered time perception had not been compared directly with objective task performance in this context. Scholle and colleagues therefore aimed to test whether psilocybin changes time perception in healthy volunteers as measured both objectively and subjectively. Specifically, they examined whether psilocybin alters performance on a visual temporal bisection task and whether any task-based changes correspond with self-reported changes in the passage of time on psychedelic phenomenology scales. The study is presented as an exploratory analysis nested within a larger clinical trial, with the authors emphasising that it extends earlier work by using a different timing task and visual rather than auditory stimuli.

This was a double-blind, placebo-controlled, within-subject study in healthy volunteers. The experiments were part of a larger clinical trial approved by the relevant ethics and drug regulatory bodies, and the article states that the timing analyses were exploratory and the study was not pre-registered. Participants received both psilocybin and placebo in separate sessions, with order determined by block permutation randomisation. The researchers report that blinding was not formally assessed. Sessions were conducted by a consistent mixed-gender team, and the same intake procedures were used across both conditions. A total of 25 healthy adults were recruited by snowball sampling; one participant was excluded because of corrupted data, leaving 24 participants for analysis. Participants were aged 28 to 55 years. Inclusion criteria required no psychiatric history and no family history of psychotic disorder up to second-degree relatives. Exclusion criteria included pregnancy, hypertension, stroke, congestive heart failure, pacemaker or cranial metal, celiac disease, left-handedness, daily medication use other than contraception, and being a medical student. Participants were screened with the Mini-International Neuropsychiatric Interview and the Minnesota Multiphasic Personality Inventory-2, and were asked to abstain from alcohol and other psychoactive substances beforehand. Psilocybin was given orally in capsules prepared under Good Manufacturing Practice conditions. The target dose was 0.26 mg/kg, achieved using combinations of 1 mg and 5 mg capsules. Actual doses ranged from 16 to 24 mg, with a mean of 19 mg. Placebo sessions used the same number of capsules. The temporal bisection task was administered 260 minutes after dosing, during the descending phase of intoxication, when the authors expected the strongest perceptual disturbances to have eased enough for the task to remain feasible. Visual stimuli were used to reduce possible motor confounding, and a simple visual hallucination task was included to check whether participants could still perceive simple images adequately at that point. Blood samples were taken repeatedly and psilocin concentrations were measured by liquid chromatography-mass spectrometry. The main objective measure was the temporal bisection task. Participants learned to discriminate a short standard duration of 1000 ms and a long standard duration of 3000 ms, then completed 56 test trials involving five durations from 1000 ms to 3000 ms. Responses were modelled with binomial logistic regression to derive two indices: the bisection point (BP), which estimates the point at which a stimulus is judged equally likely to be short or long and therefore reflects timing bias, and the just noticeable difference (JND), which reflects temporal precision, with smaller values indicating better discrimination. Subjective time experience was assessed using selected items from the Altered States of Consciousness questionnaire and the Hallucinogen Rating Scale, alongside the Brief Psychiatric Rating Scale. The researchers analysed the data in R using linear mixed models, with stepwise elimination guided by Bayesian information criterion, and they used multiple imputation for a small amount of missing data. They also used Spearman correlations with Holm correction for exploratory associations between measures.

Under psilocybin, both objective timing measures differed from placebo in the initial paired comparisons. The bisection point shifted rightwards, with t(23) = 2.27, p = 0.033 and a small effect size (g = -0.37), indicating that participants tended to judge longer durations as matching the reference midpoint, consistent with subjective time slowing. JND was also higher under psilocybin, with t(23) = 2.48, p = 0.021 and g = -0.47, indicating reduced temporal precision and a shallower psychometric function. The authors note that these effects were most evident for durations longer than 2 seconds. In the mixed-model analyses, condition remained the only significant predictor in the final models for both outcomes. For JND, ConditionPSI was the only surviving predictor and remained significant; for BP, ConditionPSI likewise remained the only predictor in the best-fitting model. Bootstrapping with 5000 resamples suggested that the BP estimate was reliable, whereas the bootstrapped confidence interval for the JND estimate included zero, indicating that the JND finding was less robust than the original model suggested. Exploratory correlational analyses did not identify strong associations between the objective measures and the subjective time items. The item ASC_Q133, which asked whether time was passing much slower than usual, loaded most strongly in a factor analysis of selected time-perception items and was significantly higher under psilocybin than placebo, t(23) = 5.72, p < 0.001. However, ASC_Q133 did not significantly predict either BP or JND. Within the psilocybin condition, ASC_Q133 was not correlated with BP, JND, or the correlation between BP and JND. By contrast, the HRS item measuring the perceived intensity of time alteration did not predict JND, but it did significantly predict BP in the psilocybin condition, alongside age. The authors also report that the simple visual hallucination check suggested that by the time of the bisection task, participants had regained sufficient perception of simple objects to complete the task, and no one was excluded on this basis.

The authors interpret the findings as evidence that psilocybin alters time perception both objectively and subjectively. They argue that the rightward shift in BP reflects underestimation of elapsed intervals, while the larger JND suggests less precise temporal judgements. They also see the increase in subjective reports of time slowing as broadly consistent with the task-based findings, particularly the BP shift. In their view, the pattern fits a Bayesian framing of timing, but they also discuss the results using Scalar Expectancy Theory, where the BP shift could be read as a slowing of the internal clock. They compare their findings with earlier human psilocybin studies, noting that the interval range at which effects emerged was similar to previous work showing impairments for durations above roughly 2-3 seconds. They suggest that the apparent threshold may be task dependent, but broadly consistent across paradigms. The authors also point out a tension with earlier psilocybin studies that reported under-reproduction of intervals, which might imply faster internal timing under a simple pacemaker-accumulator account. To reconcile this, they propose that psilocybin may affect memory, attention and other higher-order cognitive processes rather than directly changing clock speed, or that different timing tasks impose different cognitive and motor demands. They discuss Bayesian accounts, including the REBUS framework and broader ideas about altered priors and noise, but conclude cautiously that the data are more easily explained by disrupted cognition, especially for longer intervals. The authors acknowledge several limitations. The sample was small and sized mainly for the broader fMRI study rather than the timing analyses, reducing power for subtle effects. The subjective time questions were asked retrospectively at the end of the session, which introduces memory bias and makes them poorly time-aligned with the bisection task. They also note that using isolated questionnaire items is not standard practice and that their psychometric properties are unknown, although they tried to compensate by comparing multiple items and using factor analysis. The timing task was done during the descending phase of psilocybin effects, so effects may have differed if tested at peak intoxication. They also note the lack of formal blinding checks and that they did not verify whether participants complied with the instruction not to count. The authors suggest that future studies should use more precise subjective time measures, explore neural mechanisms, and consider receptor antagonists such as 5-HT2A and 5-HT1A blockers to disentangle serotonergic contributions to timing changes.

The authors conclude that psilocybin affects time perception in humans, both as a subjective experience and as measured by a temporal bisection task. They state that the effect appears to generalise across sensory modality and task type, support a threshold effect around 2-3 seconds, and suggest that the observed distortions are more likely driven by psilocybin-related cognitive disruption than by a direct change in the speed of an internal clock.