Alcohol Use Disorder (AUD)

This report summarises what Blossom’s database shows about psychedelic and dissociative compounds for alcohol use disorder, and what it does not show. The short version: alcohol is the strongest case in the addiction field. It has the single best result, a positive double-blind psilocybin trial, plus genuine ketamine evidence and famous historical LSD data. But it is also a field where the most recent, better-controlled trials have been more sobering, and the honest position is real, leading promise that still awaits confirmation.

A note before the evidence

This page is a research summary, not medical advice, and nothing here is a treatment recommendation. None of the compounds discussed is approved for alcohol use disorder; they are investigational and studied only in clinical trials with intensive psychological support and medical screening. Alcohol use disorder is a serious, treatable condition, and effective help, including approved medicines and psychotherapy, is available now through health services. Withdrawal from heavy drinking can be medically dangerous and should never be attempted with a psychedelic; decisions about treatment belong with a qualified clinician.

This page sits within Blossom’s Substance Use Disorders hub and is scoped to alcohol specifically. The shared mechanisms and the evidence in tobacco, opioids and stimulants are covered there; here the focus is the alcohol-specific trials. A word on the numbers: Blossom tracks 113 papers and 46 trials tagged to this topic, and those counts appear on the page. The tag is leaky, pulling in general-addiction reviews and off-target studies (tobacco, opioids, PTSD, depression), so the genuinely alcohol-specific clinical core is smaller, roughly two dozen human studies. Read the counts as breadth of coverage, not as the size of the alcohol evidence base.

Psilocybin: the field’s strongest single result

The reason alcohol stands out is one trial. In 2022, a double-blind randomised controlled trial (95 randomised, 93 analysed) compared two psilocybin sessions with an active placebo (diphenhydramine), both wrapped in 12 weeks of psychotherapy, and found that psilocybin reduced the percentage of heavy drinking days to 9.7%, against 23.6% for placebo, a difference of 13.9% over the 32-week follow-up[1]JAMA Psychiatry, Bogenschutz RCT (2022). That is the first positive double-blind trial of a classic psychedelic for any addiction, and it followed an open-label pilot in which abstinence rose specifically after psilocybin was given[2]J. Psychopharmacology, psilocybin pilot (2015).

It deserves to be taken seriously, and it deserves its asterisks. It is a single Phase 2 trial at two centres with around ninety patients, and 2025 brought two more double-blind psilocybin alcohol trials that did not simply confirm it. A Swiss trial of a single dose given after withdrawal found no benefit over placebo[3]eClinicalMedicine, relapse-prevention RCT (2025), and a French feasibility trial in patients with co-occurring depression reported higher abstinence (55% versus 11%) but acknowledged its blinding had failed for over 90% of participants[4]Addiction, comorbid-depression RCT (2025). A separate open-label single-dose study did report large reductions in heavy drinking[5]J. Psychopharmacology, single-dose pilot (2025), but without a control group. The pattern suggests the effect is real but sensitive to dose, number of sessions, the therapy and the population, which is exactly why replication matters.

Ketamine: a real, more modest signal

Ketamine, a dissociative anaesthetic rather than a classic psychedelic, has the second-strongest evidence. The KARE trial (n=96) found that three ketamine infusions plus relapse-prevention therapy produced significantly more days abstinent at six months than placebo, a difference of about 10%, and about 16% for the full ketamine-plus-therapy combination versus saline plus education[6]Am. J. Psychiatry, KARE ketamine trial (2022). An earlier pilot (n=40) found a single infusion increased abstinence and delayed relapse[7]Am. J. Psychiatry, ketamine pilot (2020).

The honest qualifications are that KARE improved days abstinent but not the relapse rate, the samples are proof-of-concept size, and ketamine acts briefly, so it implies a repeated, supervised, therapy-paired model rather than a single cure. A telling detail connects it to the classic psychedelics: a reanalysis found ketamine’s benefit on drinking was carried by the intensity of the mystical-type experience, not by dissociation[8]J. Psychopharmacology, mystical mediation (2020). Whatever the receptor pharmacology, the psychological character of the session seems to matter across these very different drugs.

LSD: famous history, dated evidence

LSD’s link to alcoholism is the oldest story in the field; in the 1950s and 1960s it was one of LSD’s main psychiatric uses. The modern summary is a 2012 meta-analysis that pooled six randomised trials from that era (536 patients) and found a single LSD dose reduced alcohol misuse, with an odds ratio of 1.96 and remarkably consistent results across trials[9]J. Psychopharmacology, LSD meta-analysis (2012). For a treatment given decades before modern trial standards, that is a striking and genuine signal.

But it must be read in context. Those trials used single doses, often imperfect blinding and inconsistent outcome measures; a contemporary re-analysis judged four of the six at high risk of bias[10]Current Medical Research & Opinion review (2023), and the benefit in the original meta-analysis had faded by one year. A modern, properly controlled LSD alcohol trial is now under way but has not reported. So the LSD evidence is historically important and points the same way as the newer work, without itself meeting today’s bar for proof.

MDMA, 5-MeO-DMT and others: early signals

The remaining compounds are earlier still. MDMA-assisted therapy for alcohol use disorder rests on an open-label safety study (n=14) in which drinking was much lower nine months after detox[11]J. Psychopharmacology, MDMA pilot (2021), encouraging but uncontrolled; its more promising use may be where alcohol problems and post-traumatic stress overlap. Inhaled 5-MeO-DMT reported a striking open-label phase 2a result, with abstinent days rising from 33% to 81% by week 12[12]Addiction, 5-MeO-DMT phase 2a (2025), again without a control group. Ayahuasca and ibogaine appear mainly in preclinical and observational work for alcohol, with no controlled human trial in the database.

The common thread is that beyond psilocybin, ketamine and the historical LSD data, the alcohol evidence is open-label, observational or preclinical. Large open-label effects are the norm in this field and routinely shrink under blinding, so these should be read as reasons to run controlled trials, not as results in their own right.

How might these work?

The proposed mechanisms are plausible and still partly hypotheses. The classic psychedelics act on the brain’s 5-HT2A serotonin receptors and appear to open a window of heightened neuroplasticity, which, paired with psychotherapy, may help loosen entrenched habits and the rumination and hopelessness that sustain heavy drinking. Ketamine works through a different receptor (blocking NMDA) but seems to reach a similar place, a burst of plasticity and a meaningful psychological experience. Across both, reduced craving, greater psychological flexibility and the subjective depth of the session keep emerging as the active ingredients, which is why the therapy around the drug is treated as part of the treatment, not an optional extra.

Durability, blinding and the psychotherapy question

Three practical issues run under all of this and shape how the results should be read. The first is durability: the LSD meta-analysis benefit faded by a year, ketamine acts only briefly, and even psilocybin’s single-dose model has not been followed long enough across trials to know how long the effect lasts or whether it needs repeating. The second is blinding. A high dose of a psychedelic is unmistakable, so participants usually know which arm they are in, and the 2025 French trial showed this starkly with blinding failing for over ninety per cent of patients. That expectancy almost certainly inflates the apparent drug effect, and it is the central methodological problem the field has yet to solve.

The third is the psychotherapy itself. In every positive trial the drug is given inside an intensive course of structured therapy, with preparation and integration sessions, and the largest ketamine effect was specifically the combination of drug plus therapy. It is genuinely unresolved how much of the benefit is pharmacological and how much is the psychological scaffolding, and that question matters enormously for how, and whether, any of this could be delivered at scale outside a research setting. None of this erases the signal, but it does mean the honest unit of treatment here is "drug plus therapy in a supervised container", not a pill.

Reading this honestly

So where does alcohol use disorder sit? At the front of the addiction field, and that lead is real: it has the one positive double-blind psychedelic addiction trial, a genuine ketamine signal, and a deep historical literature all pointing the same way. It is also a cautionary tale in progress. The most rigorous recent trials have been mixed, blinding is a serious unsolved problem, the samples are small, and the psychotherapy is doing a great deal of the work. The honest conclusion refuses both the hype that treats alcohol as solved and the cynicism that dismisses a striking, replicated-in-spirit body of evidence. For a condition that harms hundreds of millions of people and is poorly served by existing treatments, a leading but unconfirmed signal is genuinely hopeful, and the confirmatory trials now running will decide how much of that hope holds.