Why Andrew Huberman Changed His Mind About Psychedelics

In conversation with podcast host Tim Ferriss, Stanford neuroscientist Andrew Huberman explains how his views on MDMA and psilocybin shifted after following MAPS-sponsored trials and completing physician-guided MDMA sessions of his own for trauma. He traces the shift to a mainstream Stanford Magazine feature, a retracted MDMA neurotoxicity claim, and psilocybin trial data for treatment-resistant depression. He frames this as one scientist's account of developing research, not medical advice.

A Stanford alumni magazine normalizes the topic

Huberman describes a recent Stanford Magazine issue covering classic psychedelics (psilocybin, mescaline, LSD), ketamine, and MDMA, including their chemistry, history, clinical trials, benefits, and risks. He says a conversation like this could have ended his career even five years earlier. He credits MAPS and labs including Matthew Johnson's, Robin Carhart-Harris's, and Roland Griffiths's, plus public education from Michael Pollan, and notes MDMA is moving toward legalization for clinician-administered use in the US, with psilocybin likely further behind.

From regretted recreational use to guided therapy

He is direct about his own history: recreational LSD trips in high school lasting 11 to 15 hours, taken without knowing the sourcing, which he calls a "terrible idea" he regrets. Recreational psilocybin as an adult, he says, did little for him. He has since completed multiple MAPS-affiliated, physician-guided MDMA sessions for trauma as part of a study, calling them immensely beneficial, while stressing this differs sharply from unsupervised use, a distinction central to harm reduction.

Reassessing the safety and depression data

Huberman says he was taught MDMA was neurotoxic, based on a Science magazine study later found to have tested methamphetamine, not MDMA; the retraction drew far less attention than the original claim. On psilocybin, he cites Matthew Johnson's and Robin Carhart-Harris's work with treatment-resistant depression, where roughly 25 milligrams of synthetic psilocybin produced substantial, ongoing relief in 60 to 70 percent of patients who had not responded to talk therapy, antidepressants, or TMS. He notes pharmaceutical efforts to strip out the hallucinogenic component, raising an open question, outside his expertise, about whether the subjective experience itself drives the antidepressant effect.

Inside a guided session, and a field going mainstream

Two elements of these sessions, he says, were considered essential: an eye mask worn throughout, directing attention inward rather than to the external environment, echoing the broader emphasis on set and setting; and a deliberately sequenced music program moving from distant percussion to emotionally activating instrumentals. He frames the dosing session together with the integration work that follows as central to why guided results differ from recreational use, and contrasts today's openness with the 1960s and 70s, when Harvard and Stanford professors were pushed out for similar discussions.

Key takeaways

• ●Huberman says his public stance on MDMA and psilocybin changed markedly over roughly two years, driven by trial data and his own guided therapy.
• ●A Stanford Magazine feature covering psilocybin, mescaline, LSD, ketamine, and MDMA read to him as a sign of mainstream normalization.
• ●He regrets his own unsupervised, unsourced LSD use in high school and does not recommend it.
• ●He has completed multiple MAPS-affiliated, physician-guided MDMA sessions for trauma and calls them immensely beneficial.
• ●A widely taught MDMA neurotoxicity claim traced to a study that had tested methamphetamine, not MDMA, and was retracted with far less attention.
• ●In trials on treatment-resistant depression, he cites 60 to 70 percent of patients showing substantial, ongoing relief after guided psilocybin sessions.
• ●Guided sessions pair a full-session eye mask with a sequenced music program, and whether the subjective experience itself drives the antidepressant effect remains an open question.