Substance Use Disorders (SUD)

This report summarises what Blossom’s database shows about psychedelic treatments for substance use disorders (SUDs), and what it does not show. The short version: addiction is one of the oldest and broadest hopes for psychedelic medicine, and psilocybin now has positive randomised trials across alcohol, tobacco and cocaine, which is a genuinely unusual breadth. But the evidence is mixed rather than settled: a major alcohol trial was null, the studies are small and short, opioids are barely studied, and almost every trial struggles to keep patients unaware of whether they received the drug. Both the promise and the fragility are real.

A note before the evidence

This page is a research summary, not medical advice, and nothing here is a treatment recommendation. No psychedelic is an approved treatment for any substance use disorder. Addiction is treatable, and effective, evidence-based options exist (including life-saving opioid-agonist treatment); decisions about care belong with a qualified clinician. If you are struggling with substance use, please seek professional support.

This is the hub page. Blossom tracks 573 papers and 109 trials under this topic, and those counts appear on the page, but the tag is broad, spanning every substance plus a lot of mechanism and comorbidity work, so the genuinely outcome-focused core is smaller. Because addiction is a family of distinct disorders, our most-studied substances, alcohol, opioids and tobacco, have their own dedicated pages that go deeper; this page covers the cross-substance picture. Read the counts as database coverage, not as a tally of proven treatments. Where a key fact sat outside the database (such as the original Johns Hopkins smoking pilot or the historical LSD meta-analysis), it has been verified against primary sources and cited as such.

A long history, and why addiction is the original target

Addiction was where psychedelic therapy began. In the 1950s and 1960s, LSD was trialled extensively for alcoholism, and a 2012 meta-analysis that pooled six of those randomised trials (536 patients) found a single dose produced a modest but real reduction in alcohol misuse (odds ratio 1.96)[1]J. Psychopharmacology, LSD-for-alcoholism meta-analysis (2012). The modern era has revived that idea with better tools and, importantly, has broadened it well beyond alcohol. The underlying logic is that addiction is a disorder of rigid, over-learned reward and habit circuits, and that a psychedelic experience plus therapy might loosen those patterns in a way daily medication cannot.

Psilocybin: the workhorse with the broadest case

Psilocybin is the most-studied compound here and the only one with controlled trials across several substances. In alcohol, the 2022 NYU trial found that psilocybin plus psychotherapy roughly halved the percentage of heavy drinking days compared with an active placebo[2]JAMA Psychiatry, Bogenschutz psilocybin AUD RCT (2022). In tobacco, a Johns Hopkins trial reported that 40.5% of psilocybin participants achieved verified abstinence at six months versus 10.0% on a nicotine patch[3]JAMA Network Open, psilocybin vs nicotine patch (2026), building on an earlier open-label pilot in which 80% were abstinent at six months[4]J. Psychopharmacology, Johns Hopkins smoking pilot (2014). And in 2026 a quadruple-blind trial reported a large increase in cocaine-abstinent days[5]JAMA Network Open, psilocybin cocaine RCT (2026), notable both for its result and its unusually rigorous design.

The qualifications matter as much as the wins. The alcohol evidence is genuinely contradictory: a 2025 Swiss Phase 2 relapse-prevention trial found no benefit over placebo at all[6]eClinicalMedicine, psilocybin AUD relapse-prevention (2025), while a positive French trial in alcohol use disorder with depression saw 55% abstinence versus 11% but had 93% of patients correctly guess their assignment[7]Addiction, psilocybin AUD + depression RCT (2025). A small methamphetamine pilot was encouraging but open-label[8]Addiction, psilocybin methamphetamine pilot (2025). So psilocybin is best described as a broad, real signal that is strongest in tobacco and cocaine, mixed in alcohol, and everywhere limited by small samples and imperfect blinding.

Alcohol: the deepest but most contradictory evidence

Alcohol has the most psychedelic research of any substance, and also the most conflicting. Alongside the positive Bogenschutz trial and the null Swiss trial, ketamine has a place: the KARE trial found three ketamine infusions plus therapy increased abstinence in severe alcohol use disorder[9]Am. J. Psychiatry, KARE ketamine AUD trial (2022), and an early 5-MeO-DMT proof-of-concept reported abstinent days rising from 33% to 81%[10]Addiction, 5-MeO-DMT (BPL-003) AUD (2025) in a dozen patients. The breadth is encouraging, but the contradictions are the headline. For the deeper picture, see our dedicated alcohol use disorder page.

Tobacco: the cleanest single signal

Tobacco produced one of the most striking results in the whole field. The Johns Hopkins randomised trial found psilocybin roughly quadrupled six-month abstinence compared with a nicotine patch[3]JAMA Network Open, psilocybin vs nicotine patch (2026). The caveat is honesty about blinding (the trial was openly unblinded) and scale (a pilot of 82 people), and a separate ketamine smoking trial was null[11]Brain Sciences, ketamine tobacco crossover (2026). Still, for a single, well-defined outcome, the psilocybin smoking signal is among the cleanest the field has. Our tobacco use disorder page covers it in detail.

Stimulants: a genuinely new result

Cocaine and methamphetamine are important because there are no approved medications for them at all, so any signal is meaningful. The standout is the 2026 quadruple-blind psilocybin cocaine trial, which found markedly more cocaine-abstinent days and a lower risk of lapse[5]JAMA Network Open, psilocybin cocaine RCT (2026), in a predominantly underserved population. A methamphetamine pilot also showed reduced use[8]Addiction, psilocybin methamphetamine pilot (2025), though open-label. These are early but, given the complete absence of approved options, among the most consequential findings here.

Opioids and ibogaine: promise shadowed by real risk

Opioids are the thinnest area, and the most cautionary. The psychedelic most associated with opioid addiction is ibogaine, reported, largely through clinics and case studies, to produce dramatic reductions in withdrawal and craving. A 2026 case study of ten opioid-dependent people described rapid detoxification and periods of abstinence[12]J. Psychoactive Drugs, ibogaine opioid case study (2026). But there is no controlled efficacy trial, and the safety problem is severe: ibogaine prolongs the heart’s QT interval and can cause fatal arrhythmias[13]Molecules, ibogaine cardiac-safety scoping review (2026), with deaths linked to unsupervised use. This is why, for opioids, the honest message leads with caution rather than promise. Our opioid use disorder page goes further.

How psychedelics might treat addiction

The proposed mechanisms converge on flexibility. Classic psychedelics act on the brain’s 5-HT2A receptors and appear to increase neuroplasticity, loosening rigid patterns of thought and behaviour. A systematic review found that psychedelic-induced insight was associated with therapeutic improvement, often more strongly than the intensity of the mystical experience itself[14]Neurosci. Biobehav. Reviews, insight systematic review (2025), pointing to psychological change as a key driver. There is also a more direct circuit story: DMT has been shown to reduce connectivity in the midbrain-to-nucleus-accumbens reward pathway that is typically over-active in addiction[15]Scientific Reports, DMT VTA-NAc connectivity (2025). Ketamine, by contrast, works through glutamate and rapid synaptic plasticity. The common thread is interrupting the entrenched reward and habit circuitry that sustains compulsive use.

The unblinding problem, and what is still missing

One issue shadows every result on this page: it is extremely hard to run a blinded psychedelic trial when patients can feel whether they received an active drug. The French alcohol trial’s 93% correct-guess rate[7]Addiction, psilocybin AUD + depression RCT (2025) is the clearest illustration, and a major 2026 review listed functional unblinding among the central limitations of the entire field[16]BMJ state-of-the-art review (2026). Combined with small samples, short follow-ups, and the near-absence of opioid and cannabis trials, this means the field’s real task now is not more pilots but larger, better-controlled studies. Encouragingly, the most rigorous recent trial (the quadruple-blind cocaine study) is a sign that researchers are taking this seriously.

Who is doing the work

Addiction research is unusually academic. The defining trials come from universities (NYU, Johns Hopkins) and public funders rather than a single dominant company, which gives it a more investigator-led character than depression or PTSD. Commercial interest is rising, especially in psilocybin and novel tryptamines for alcohol and stimulants, and in cardiac-safer ibogaine formulations for opioids[13]Molecules, ibogaine cardiac-safety scoping review (2026). The clearest opportunity is in the stimulant and cannabis disorders, where no approved medication exists at all.

Reading this honestly

So where do substance use disorders sit? Among the more promising psychedelic indications, and among the oldest, but not among the most settled. Psilocybin’s breadth across alcohol, tobacco and cocaine is genuinely distinctive, and the stimulant results are especially valuable given the lack of any approved alternative. Yet the null alcohol trial, the unblinding, the small samples and the serious safety issues with ibogaine all argue against over-claiming. The honest verdict is a real, broad and historically grounded signal that now needs to prove itself in large, rigorously blinded trials, substance by substance. For people living with addiction, that combination, genuine hope without a proven cure, is both the most encouraging and the most truthful thing the evidence will currently support.