The mechanism question, how psychedelics actually work in the brain, is both simpler and murkier than a lot of popular coverage suggests. The simpler part is the receptor story. The murkier part is everything that follows after that first pharmacologic hit.
What we know
For classic psychedelics, the best-supported starting point is still the serotonin 5-HT2A receptor. A 2025 human imaging study comparing psilocybin with the 5-HT2A antagonist ketanserin adds fresh support to that view. In 28 healthy participants, psilocybin was associated with lower regional and global cerebral blood flow at peak effect, while ketanserin alone did not produce the same pattern. The paper argues those acute effects are consistent with 5-HT2A receptor neuromodulation in humans, not just a vague serotonin story.
That is a useful correction to the way psychedelic neuroscience is often summarized. These drugs are not just "boosting serotonin" in the broad SSRI sense. They are acting on a specific receptor system with specific downstream consequences, and the field has gotten more confident on that narrow point than on almost anything else.
What we are still figuring out
The real uncertainty is what happens after receptor activation. One of the leading ideas is that psychedelics may promote structural plasticity, meaning changes in the brain's capacity to form or strengthen connections. But the strongest direct evidence is still uneven and often preclinical.
A 2021 psilocybin study in pigs is useful here because it looked directly at SV2A, a marker used as a proxy for synaptic density. One day after a single psilocybin dose, the researchers found higher hippocampal SV2A density and lower 5-HT2A receptor density. Seven days later, SV2A density was still elevated in the hippocampus and prefrontal cortex, while the receptor changes had faded. That does not prove the same timeline holds in humans, but it is a more concrete basis for the neuroplasticity discussion than the usual hand-waving.
The timing question remains especially important. If psychedelics really do open a temporary window of heightened plasticity, clinicians would want to know whether that window is measured in hours, days, or longer. Right now, that part of the story is still early.
The imaging problem
One reason the mechanism story is hard to pin down is that some of the most striking human imaging results are also the easiest to overread. The 2025 cerebral-blood-flow paper found that psilocybin was associated not only with lower cerebral blood flow, but also with a measurable decrease in internal carotid artery diameter. In plain terms, psilocybin appears to affect the vasculature as well as the neurons.
That matters because fMRI uses blood flow as a proxy for neural activity. If psychedelics alter blood vessels directly, then some apparent brain-network changes may reflect altered neurovascular coupling, not just cleaner readouts of neuron-to-neuron signaling. This does not make the imaging literature useless, but it does mean dramatic connectivity images should be interpreted more cautiously than they often are.
Why this matters
A 2025 review on psychedelic-induced neuroplasticity makes the current frontier pretty clear: preclinical evidence for structural plasticity is now substantial, but translatable human biomarkers are still lagging. The authors argue that newer PET methods could help measure synaptic proteins in humans after psychedelic dosing, which would make it easier to identify responders, optimize dosing schedules, and test whether new compounds really reproduce the biology that people think matters.
That is why the best answer to "what do psychedelics do to the brain?" is still two-part. First, they act through 5-HT2A-mediated signaling in ways that clearly change acute brain physiology. Second, researchers are still working out whether the durable therapeutic story is really about synaptic remodeling, how to measure that in living humans, and when those changes matter most.