Set and setting isn't woo, it's protocol

"Set and setting" arrives with a whiff of incense about it, which does the idea a disservice. In clinical practice it is not a vibe. It is a checklist, and much of the safety record of modern psychedelic research rests on it.

Where the phrase comes from

It entered the research literature through Timothy Leary, who presented on the concept at the American Psychological Association meeting in September 1961. The historian Ido Hartogsohn, whose 2017 review is the definitive account, is careful to say the coining is "commonly credited" to Leary rather than demonstrated, and lists earlier antecedents: Al Hubbard's attempts to build a therapeutic environment for LSD sessions, Betty Eisner's 1957 account of incorporating both factors into therapy, Robert Hyde's work on the parameters around the drug. Norman Zinberg generalised the idea to all intoxicants in his 1984 book Drug, Set, and Setting.

Set means the person: their expectations, their preparation, their state of mind on the day. Setting means everything around them: the room, the people, the music, what happens next.

What it looks like as procedure

The foundational document is Johnson, Richards and Griffiths, "Human hallucinogen research: guidelines for safety", published in the Journal of Psychopharmacology in 2008. Its core safeguards are unglamorous and specific: screen out volunteers with a personal or family history of psychotic disorders; build trust and rapport between the participant and their session monitors before the day; prepare the volunteer properly; provide a safe physical environment; and have at least two trained monitors present throughout.

The paper describes the Johns Hopkins session room as "living room-like", deliberately "aesthetically pleasing and free of extraneous medical or research equipment." A hospital room signals emergency. A sitting room does not.

Eyeshades and headphones are standard, and the paper's stated reason is not to intensify anything but to reduce distraction and the social pressure to talk to the researchers. The music is not background: the Johns Hopkins session playlist, curated principally by Bill Richards, runs about seven hours and is mapped to the arc of a session. Participants are read what the field calls flight instructions, the guidance often summarised as trust, let go, be open.

In industry trials this is manualised. In Compass's psychological support model, participants meet a qualified therapist at least three times before dosing, sit with two trained therapists on the day with a doctor on call, and return for integration sessions including one the day after. Imperial's model has a name, Accept-Connect-Embody. None of this is decoration. It is roughly 20 hours of human contact around a single dose.

The evidence that it matters

Some of it is suggestive rather than conclusive, and we will be honest about which is which.

A 2018 study by Mendel Kaelen and colleagues in Psychopharmacology, titled "The hidden therapist", found that in 19 patients with treatment-resistant depression, the quality of the music experience predicted reductions in depression a week later, while the general intensity of the drug effect did not. Nineteen people is a very small study. It is also a striking result.

Two independent research groups have found that the therapeutic alliance, the working relationship between participant and guide, relates to outcomes. In Murphy and colleagues' 2022 analysis of 30 patients, the effect on clinical outcome was indirect, running through rapport and the quality of the acute experience.

The clearest evidence comes from the other direction. In Carbonaro and colleagues' 2016 survey of nearly 2,000 people recalling their worst mushroom experience, the factors associated with risky behaviour were dose, duration, difficulty, and the absence of physical comfort and social support. A 2023 study of people reporting extended difficulties afterwards found a narrower range of problems among those who had taken the drug in a guided setting.

And the part that must be said

A clinical setting is not automatically a safe one, and the field's own record proves it. In a MAPS-sponsored phase 2 MDMA trial in British Columbia, session video showed the therapist pair pinning and lying against a participant, and one of the therapists later admitted a sexual relationship with her while she was still enrolled. MAPS publicly acknowledged the conduct as unethical in 2019 and cut ties. When the FDA rejected the MDMA application in 2024, therapy misconduct was among the concerns cited.

Set and setting is not a promise that a room can be made safe. It is the accumulated knowledge of what has to be in place before anyone should try, plus the uncomfortable reminder that the people in the room are part of the setting too.

Carbonaro's paper puts the positive case in one sentence: "The incidence of risky behavior or enduring psychological distress is extremely low when psilocybin is given in laboratory studies to screened, prepared, and supported participants." Every word in that sentence is load-bearing.