The shape of the evidence
Compare this field to psilocybin. Psilocybin has multiple randomised controlled trials, published dose-response work, and regulatory pathways in progress. 5-MeO-DMT has a much smaller literature: a set of naturalistic and survey studies, some pharmacological and animal work, case reports, and a small number of early-phase clinical trials of synthetic material.
That is not a reason to dismiss it. It is a reason to be careful about how strongly anyone states a conclusion, including us.
Survey and naturalistic work
The most cited human data comes from surveys of people who used 5-MeO-DMT outside a clinical setting, and from observational studies conducted at retreat-type settings. These have reported improvements in depression, anxiety, and life satisfaction, along with high rates of experiences that score as mystical on standard instruments.
The limitations are structural rather than incidental. Participants are self-selected and often paid substantially to be there. There is no control group. Outcome measures are self-reported, sometimes by people with a strong investment in the answer. Follow-up is usually weeks to months. Expectancy effects in this setting are enormous. People who had a bad time are systematically less likely to still be in touch to answer a questionnaire.
Read as a signal worth investigating, this work is genuinely interesting. Read as evidence of efficacy, it is being asked to carry more than it can.
Clinical trials
Several companies have advanced synthetic 5-MeO-DMT, referred to in that literature as mebufotenin, into early clinical work for treatment-resistant depression, using inhaled and intranasal formulations. Early-phase results reported through the mid-2020s described rapid reductions in depression scores in small samples, with the usual caveats about open-label design, unblinding, and short follow-up.
Two things are worth holding at once. First, this is a real and encouraging line of research. Second, early-phase results in psychiatry frequently shrink when they meet a properly controlled trial, and a rapid, unmistakable drug effect makes blinding extremely difficult.
What is missing
- Adequately powered randomised controlled trials with credible blinding.
- Long-term follow-up beyond a few months.
- Systematic adverse event collection from non-clinical settings, where almost all use happens.
- Human pharmacokinetic data across the range of CYP2D6 phenotypes.
- Any characterisation of the active contribution of bufotenine as a metabolite.
- Compositional analysis of toad-derived material at scale.
- Denominator data. We have no idea how many people have used this and therefore no idea what the rate of anything is.
How to read a claim
Four questions handle most of it. How many people were in the study. Was there a control group. Who paid for it and who collected the outcome data. How long was the follow-up.
Apply those to any claim, including the ones on this website, and most of the confident noise in this field resolves into a much smaller number of things that are actually known.