Inside a brain on psilocybin, something unusual happens to the signal. The neurons keep firing, more or less as before, but the statistical pattern of that firing changes in a way that information theorists would recognise instantly: the data stream becomes harder to compress. Less predictable. Richer, in a technical sense, with information. This quality, which researchers call brain entropy, turns out to matter quite a lot, not just for what happens during the six or so hours a mushroom compound spends in the bloodstream, but for what happens to a person’s mind and white matter in the weeks that follow.
A study published this week in Nature Communications offers the most detailed look yet at what a single high dose of psilocybin does to the human brain, from the first hour of the experience through to a month afterward. The results are striking enough that even the researchers appear slightly surprised by some of them.
The work, led by Robin Carhart-Harris at UC San Francisco alongside colleagues at Imperial College London, recruited 28 healthy adults who had never taken a psychedelic in their lives. Each participant received two doses of psilocybin, a month apart: first 1 milligram, a sub-threshold amount that works as a functional placebo, and then 25 milligrams, a dose capable of producing a full psychedelic experience. Brain imaging came in three waves: before any dosing, a month after the placebo, and a month after the high dose. During each of the two dosing sessions, electrodes on the scalp recorded brain activity in real time.
The Brain Becomes Harder to Read
Within 60 minutes of swallowing the 25 milligram capsule, something measurable had already begun. The EEG traces showed a sharp rise in what the team measures as Lempel-Ziv complexity, essentially a calculation of how hard it is to summarise a stretch of neural signal without losing information. Alpha waves, which normally keep a kind of rhythmic order across the cortex, dropped substantially. Gamma activity climbed. The brain, in the language of information theory, had become noisier. More complex. Possibly more open.
This entropic quality peaked at around two hours post-dose, coinciding with the most intense phase of the experience. Twenty-seven of the 28 participants rated what they went through as the single most unusual state of consciousness of their entire lives; the remaining person placed it in their top five. No one who received the 1 milligram dose reported anything remotely similar.
What the team then found was a chain of prediction running forward through time. The degree to which a participant’s brain entropy had spiked during the session correlated with how much psychological insight they reported the following day, a quality assessed via a validated scale probing self-awareness and behavioural shifts. That insight, in turn, predicted improvements in well-being scores at the one-month mark. “Psychedelic means ‘psyche-revealing,’ or making the psyche visible,” said Carhart-Harris. “Our data shows that such experiences of psychological insight relate to an entropic quality of brain activity and how both are involved in causing subsequent improvements in mental health. It suggests that the trip, and its correlates in the brain, is a key component of how psychedelic therapy works.”
White Matter, Unexpectedly Changed
The stranger finding, and the one the authors are most cautious about, comes from diffusion tensor imaging, a technique that tracks how water molecules move along the brain’s white matter tracts. One month after the 25 milligram dose, participants showed decreased axial diffusivity in two bilateral pathways connecting the prefrontal cortex to subcortical structures: one running to the striatum, one to the thalamus. The same measurement taken a month after the placebo showed nothing of the sort.
Axial diffusivity measures how freely water diffuses along the principal axis of a nerve fibre. When it decreases, the tract has in some sense become denser, more organised, or structurally altered in a way that constrains that diffusion. Similar decreases have been observed after intensive meditation practice and, interestingly, after rapid learning. What they signify at the cellular level remains genuinely unclear; the change could reflect dendritic growth, altered myelination, shifts in axon density, or changes in extracellular fluid, among other possibilities. The authors flag all of this explicitly. Without replication in a larger study using more advanced imaging sequences, drawing firm conclusions about microstructural neuroplasticity would be premature.
Still, the finding rhymes with preclinical work. Studies in mice have shown psilocybin promotes rapid growth of dendritic spines in frontal cortex; pig studies have found increased synaptic density after a single dose. A possible anatomical correlate in living humans, however provisional, is new.
The DTI changes also correlated with shifts in brain network modularity, the degree to which the brain’s functional regions sort themselves into distinct, segregated clusters. After the high dose, modularity tended to decrease, meaning the brain’s networks became somewhat more globally integrated, less siloed. Participants whose modularity fell the most tended to show the greatest improvements in well-being. This same relationship has appeared in previous psilocybin trials targeting depression, though in those studies the effect was more robust; the researchers suggest that healthy brains, lacking the extreme network rigidity seen in depression, may simply have less room to shift.
What the Trip Is Actually For
Perhaps the most clinically consequential part of the analysis is the mediation model. The team tested whether the link between acute brain entropy and one-month well-being ran directly, or whether it was carried through psychological insight. Inserting insight as a mediating variable strengthened the model considerably. In plain terms: the psychedelic experience produces a specific quality of brain activity; that activity fosters self-reflection; and the self-reflection drives lasting change. The experience itself, not merely the pharmacology, appears to matter.
“Psilocybin seems to loosen up stereotyped patterns of brain activity and give people the ability to revise entrenched patterns of thought,” said Taylor Lyons, the paper’s first author. “The fact that these changes track with insight and improved well-being is especially exciting.”
For the emerging field of psychedelic medicine, this is arguably reassuring. It means that the therapeutic effect is not simply a molecular one that could be replicated by a drug lacking the subjective experience; it’s bound up with what happens to a person’s mind during the hours the compound is active. The implication, which the team is careful not to overstate, is that optimising the conditions for insight during a session could improve outcomes, and that the EEG signature of brain entropy might eventually serve as a real-time indicator of whether the dose and setting are producing the kind of neural state associated with therapeutic benefit. There is, of course, still a long way to go. But the chain of evidence, from entropy to insight to well-being to possible anatomical change, is getting harder to dismiss as coincidence.
Source: Lyons et al., Nature Communications (2026). doi:10.1038/s41467-026-71962-3
Frequently Asked Questions
Does the psychedelic experience itself matter for psilocybin’s therapeutic effects, or is it just the drug?
The experience appears to matter considerably. This study found that the degree of brain entropy during the trip predicted how much psychological insight people felt the next day, and that insight, in turn, predicted well-being improvements a month later. Stripping out the subjective experience from the pharmacology may therefore reduce therapeutic benefit, which has significant implications for how psilocybin-based treatments are designed and administered.
Is psilocybin actually changing the physical structure of the brain?
Possibly, though with important caveats. One month after a high dose, participants showed changes in white matter tracts connecting the prefrontal cortex to deeper brain structures, a pattern not seen after the placebo. The researchers caution that the specific biological cause of this change is unclear and that the finding needs replication in larger studies before any firm conclusions can be drawn. Similar patterns have appeared after meditation and intensive learning, so the change, if real, is not necessarily alarming.
Could measuring brain entropy during a session help doctors predict who will benefit?
That is one of the study’s more intriguing implications. Participants with the largest spikes in brain entropy were also the most likely to report insight and improved well-being weeks later. If that relationship holds in clinical populations, EEG-measured entropy could eventually function as a real-time indicator during therapy sessions, helping clinicians assess whether conditions are right for a beneficial experience. The idea remains speculative for now, but it points toward a more personalised approach to psychedelic medicine.
Why did the researchers use healthy volunteers rather than people with depression or anxiety?
Working with healthy participants gave the team greater freedom to conduct intensive imaging and testing without the ethical constraints that come with treating a vulnerable clinical population. It also allowed them to establish a cleaner baseline and rule out effects driven by the illness itself rather than the drug. The tradeoff is that brain changes tended to be more modest than in previous psilocybin-for-depression trials, which the researchers suggest may reflect the fact that healthy brains have less rigid network organisation to begin with.
ScienceBlog.com has no paywalls, no sponsored content, and no agenda beyond getting the science right. Every story here is written to inform, not to impress an advertiser or push a point of view.
Good science journalism takes time — reading the papers, checking the claims, finding researchers who can put findings in context. We do that work because we think it matters.
If you find this site useful, consider supporting it with a donation. Even a few dollars a month helps keep the coverage independent and free for everyone.