Psilocybin for Alcohol and Addiction: What the Research Shows

Addiction is where psychedelic research started, and it’s where some of the most striking numbers in the field come from. It’s also where the evidence is messier than the headlines suggest. Several careful trials found clear benefits, and at least one careful trial found nothing.

Below is where the research stands as of fall 2026: what the meta-analyses say, which trials they rest on, and what none of it tells you yet.

The short answer

For alcohol use disorder, the fairest summary is: it probably helps, and we can’t be sure yet. The newest meta-analysis, published in October 2026, found that people who received psilocybin drank less than controls in the months after treatment and had roughly 2.8 times the odds of staying abstinent. But it rests on only three randomized trials, the effects faded on most measures over longer follow-up, and an independent review that applied stricter rules found the same direction of benefit but couldn’t rule out no effect, rating the certainty very low.

For smoking, the results are the most dramatic in the field, but they come from Johns Hopkins alone and from studies where everyone knew who got psilocybin.

For cocaine, a single randomized trial published in 2026 showed large effects. It’s promising, but it’s one study of 40 people.

All of these studies tested psilocybin inside a structured addiction-treatment program, usually with weeks of cognitive-behavioral or motivational therapy around the doses. None of them tested a psilocybin session on its own.

Alcohol: what the meta-analyses say

The newest pooled estimate. A systematic review and meta-analysis in the Journal of Psychopharmacology (October 2026) pooled six contemporary trials with 202 participants, three of them randomized and controlled. Compared with controls, psilocybin produced moderate short-term reductions in drinks per day, drinking days, and heavy-drinking days. At longer follow-up (up to 28 weeks), only the reduction in drinking days was still statistically significant. People who got psilocybin had 2.78 times the odds of maintaining abstinence. Side effects were mostly mild and short-lived. The authors’ own summary is that the evidence is “mixed” with “preliminary signals of efficacy.” That’s an accurate description.

A stricter, living review. A living systematic review in European Neuropsychopharmacology (2025) included only double-blind randomized trials, pooling psilocybin and LSD studies together. On abstinence in alcohol use disorder, the result leaned toward benefit but did not reach significance (risk ratio 1.42, confidence interval 0.89–2.26), and the authors rated the certainty very low. The trials in it agreed closely with one another; the problem is that there are few of them, they are small, and most participants could tell what they received. Read side by side, the two reviews don’t contradict each other. Both point toward a benefit. The looser review is confident enough to call it significant, and the stricter one isn’t. Together they amount to: probably helps, can’t be sure yet.

The older LSD literature. A 2012 meta-analysis of six randomized trials of LSD for alcoholism from the 1960s and ’70s (536 participants) found that a single dose roughly doubled the odds of improvement (odds ratio 1.96), with no meaningful disagreement between trials. A 2024 re-analysis reached a similar estimate but flagged that four of the six trials had a high risk of bias. Those studies are a big reason the modern trials exist, but they used methods that wouldn’t pass review today. For the background, see our brief history of psychedelic therapy.

Alcohol: the trials behind the numbers

The pooled estimates are only as good as the handful of trials they combine, so the trials are worth knowing individually.

NYU and University of New Mexico, 2022 — the landmark positive trial. In a double-blind randomized trial in JAMA Psychiatry, 95 adults with alcohol dependence received 12 weeks of motivational and cognitive-behavioral therapy, plus two day-long sessions with either psilocybin or an active placebo (diphenhydramine). Over the following 32 weeks, the psilocybin group had heavy-drinking days on 9.7% of days, compared with 23.6% for placebo. That is a substantial difference, and there were no serious adverse events from psilocybin. A later analysis of the same trial found lasting decreases in neuroticism and impulsiveness, and less impulsiveness went with less drinking.

Zurich, 2025 — the negative trial. A phase 2 randomized trial in eClinicalMedicine gave 37 people who had just finished alcohol withdrawal treatment either one 25mg dose of psilocybin or a placebo, along with brief psychotherapy. There was no meaningful difference between the groups in days abstinent or in how much people drank, at four weeks or at six months. Craving dropped in both groups, slightly more with psilocybin. The authors concluded that one dose with five therapy sessions “may not be sufficient” for people with severe alcohol use disorder after detox.

France, 2025 — alcohol plus depression. A small double-blind pilot in Addiction enrolled 30 people with severe alcohol use disorder and persistent depression who were in an inpatient program after detox. They received two doses of either 25mg or 1mg of psilocybin on top of standard care. At 12 weeks, 55% of the 25mg group were abstinent, compared with 11% of the 1mg group. The study was built to test feasibility, not efficacy, and nearly everyone correctly guessed which dose they had received.

Denmark, 2025 — one dose, open label. A 10-person open-label study of a single 25mg dose found that heavy-drinking days fell by about 37 percentage points over 12 weeks. With no control group, it shows feasibility, not efficacy.

The pattern across these trials is consistent: the positive results come from protocols with two doses and extended therapy, and the clearest negative result comes from one dose and brief therapy. An August 2026 systematic review of every psilocybin addiction trial reached the same conclusion. Repeated high-dose sessions combined with longer psychotherapy “consistently reported sustained clinical benefits,” and the authors argue the treatment should be understood as a complex intervention, not a pill.

Smoking

Smoking has the most striking numbers in the field, and they all come from Johns Hopkins.

In the original open-label pilot, 15 smokers received two or three psilocybin sessions alongside a cognitive-behavioral quit-smoking program. Twelve months later, 67% were confirmed abstinent. At an average of 30 months, 60% still were. For comparison, the best standard smoking-cessation treatments usually help roughly a quarter to a third of people stay quit at six months.

In March 2026, the same group published a randomized pilot in JAMA Network Open comparing one high dose of psilocybin with a standard 8–10 week nicotine patch course. Both groups got the same 13-week CBT program. At six months, 40.5% of the psilocybin group were biochemically confirmed abstinent, compared with 10% of the patch group.

Two caveats. First, the trial was unblinded: participants knew what they received, and people who signed up for a psilocybin study and got the patch may have been disappointed in ways that affected their effort to quit. Second, it hasn’t yet been reproduced by a team outside Hopkins. It’s a strong signal that still needs confirming.

Cocaine and other drugs

In May 2026, researchers at the University of Alabama at Birmingham published the first randomized trial of psilocybin for cocaine use disorder in JAMA Network Open. Forty participants received a single dose of psilocybin or active placebo, with cognitive-behavioral therapy before and after. Over six months, the psilocybin group had about 29 percentage points more cocaine-free days, a much higher likelihood of complete abstinence, and roughly a quarter of the risk of relapse at any given point. Use was confirmed by urine testing. No serious adverse events occurred. The study is also notable for who took part: most participants were Black and lower-income, groups that are badly underrepresented in psychedelic research.

That is one trial, and the confidence interval on the complete-abstinence finding is extremely wide. Methamphetamine and opioid use disorders have only small pilot data, and there’s not enough yet to say anything useful.

Who it seems to help

Averages hide the real story. In nearly every trial, some participants stopped drinking or smoking entirely and others didn’t change at all. The more useful question is what separated them, and the research has early answers, mostly about what happened during and after the session rather than about who people were going in.

What the session was like. In the first alcohol pilot, the strength of the first psilocybin session strongly predicted how much drinking dropped over the following month, along with craving and confidence about staying sober. In the Hopkins smoking pilot, the 12 people who quit scored higher on mystical-type experience than the 3 who didn’t, even though the overall intensity of the drug effects was the same. The kind of experience seemed to matter more than how strong it was. That link is real but weaker for addiction than for depression: a 2025 meta-analysis put the correlation at about 0.19 for addiction versus 0.41 for mood disorders. (More in our post on the mystical experience.)

What people did with it. The most telling finding comes from the French trial. Researchers analyzed what participants said in their integration session the day after dosing. People who were still abstinent three months later talked about inner dialogue: revisiting their drinking triggers with new ways of coping, stepping back from old patterns, committing to change, working on the intentions they had set about alcohol, and reconnecting with family. People who relapsed mostly described the sensory side of the experience (visuals, physical sensations, pain) along with disappointment, and they tended to push difficult feelings away rather than work through them.

What changed afterward. In the NYU trial, participants who became less impulsive drank less, and the effect was strongest among psilocybin participants who were still drinking heavily right before their first dose. The authors are careful to say they can’t yet tell whether psilocybin reduced impulsiveness or whether impulsive people simply respond better.

What we don’t know. No trial has been large or varied enough to show which kinds of people respond best, by severity, age, sex, treatment history, or other psychiatric conditions. Trials screen heavily, which narrows the differences there are to study. One small hint: in the French trial, people taking antidepressants did as well as people who weren’t.

Taken together, the early evidence points less to a type of person and more to a type of process. The people who improved tended to arrive with clear intentions about their drinking, engage actively with the experience, and turn it toward the problem afterward. That is the strongest argument in the research for taking preparation and integration seriously.

Why these studies are hard to read

Nearly everyone can tell what they got. A 2026 review in JAMA Psychiatry found that in psilocybin trials, participants and raters frequently guessed the treatment correctly more than 90% of the time. When people know they got the real drug, expectation can drive part of the result. Active placebos like diphenhydramine help, but not much.

The trials are small. The newest alcohol meta-analysis covers 202 people in total. One or two trials with different results can change a pooled estimate significantly, which is exactly what happened between the 2025 and 2026 reviews.

Participants are carefully selected. Most trials excluded people with other drug use disorders, serious psychiatric conditions, or significant medical illness. The NYU trial excluded anyone already in treatment for alcohol use disorder. Results from screened research volunteers don’t automatically transfer to people with more complicated histories.

The therapy is part of the treatment. Every positive trial wrapped the doses in weeks of structured therapy. We can’t yet separate what psilocybin contributes from what the therapy contributes, and the negative Zurich trial suggests the therapy is not optional.

What this means in Colorado

Colorado’s natural medicine program is not the treatment that was studied in these trials, and it’s important to be precise about the difference.

The trials were medical research: psilocybin given as part of a manualized addiction-treatment program, by research teams that included physicians, with screening, monitoring, and weeks of therapy built around each dose. Under the Natural Medicine Health Act, a licensed facilitator provides preparation, a supervised session, and integration. Facilitators are not prescribers, and a facilitated session is not, by itself, treatment for alcohol or drug use disorder.

What that means in practice:

Medically supervised withdrawal comes first. If you drink heavily every day, or depend on benzodiazepines or opioids, stopping can be dangerous. Withdrawal needs medical management before anyone schedules a session. Ongoing heavy alcohol or stimulant use is one of the situations covered in psilocybin contraindications, and it gets a real conversation during screening.

It works best alongside treatment, not instead of it. FDA-approved medications such as naltrexone and acamprosate have far more evidence behind them than psilocybin does, and they’re prescribed to only a small fraction of the people who could benefit from them. If you’re considering a session for drinking, the research model that worked was psilocybin added to an addiction-treatment plan. See how psilocybin fits with other treatments.

People in recovery differ on this. Some people in abstinence-based recovery see any psychedelic use as incompatible with their sobriety, and others don’t. Programs generally treat this as the participant’s call to make with their sponsor, therapist, or recovery community, not something a facilitator should decide for them.

Integration is where the change has to stick. Across the alcohol trials, the protocols that worked put far more structure after the dose than before it. In a session-based model, that means taking integration seriously as part of the service rather than an optional extra.

The bottom line

The research on psilocybin for addiction is promising and still early. For alcohol, psilocybin given within a structured therapy program probably helps, and we can’t be sure yet. Every review points the same way, but the evidence rests on a handful of small trials, and the most rigorous review can’t rule out no effect. For smoking and cocaine, single research groups have reported large effects that haven’t yet been replicated.

The two most consistent findings are that psilocybin was generally well tolerated in screened participants, and that the therapy around the dose appears to carry much of the weight.

---

Considering natural medicine and have a history with alcohol or other substances? Get in touch or book a free 15-minute consultation. We’ll talk through your situation, including whether medical or addiction care should come first. You may also want to read is psilocybin therapy right for you or learn about our natural medicine services.

This post is educational and is not medical advice. Psilocybin sessions under Colorado’s Natural Medicine Health Act are not a substitute for treatment of a substance use disorder. If you are physically dependent on alcohol, benzodiazepines, or opioids, talk to a medical provider before stopping.