Concord Prison Experiment (historical)

1. Introduction

Version followed: the Harvard programme at Massachusetts Correctional Institution (MCI) Concord, 1961–63, as published by Leary, Metzner, Presnell, Weil, Schwitzgebel and Kinne (1965) and as reconstructed in Rick Doblin's 34-year follow-up (1998) and Ralph Metzner's reply (MAPS Bulletin, 1999/2000). Sources: Doblin 1998 is used in full from two legitimate copies (the Journal of Psychoactive Drugs 30(4), 419–426 text on the MAPS bibliography server and the MAPS Bulletin reprint with Metzner's reply). The US and UK legal status, the 1962–63 Harvard events (contemporaneous Harvard Crimson reporting) and the named rehabilitation alternatives now have sources. The 1965 paper itself is known here only through its abstract, so programme details come from Doblin's account of it [uncertain]. The date of the first Crimson report and the 1961 student hospitalisations also rest on Wikipedia alone [uncertain where noted].

HISTORICAL, NOT FOR USE. This is not a therapy a clinician can or should deliver. It was a research programme, now known mainly as a cautionary case in research ethics and outcome reporting. No doses, administration details or session scripts are given here.

What it was: a Harvard team led by Timothy Leary (with graduate students Ralph Metzner and Gunther Weil, others including Alan Cohen, Alfred Alschuler, George Litwin, Ralph Schwitzgebel and Michael Hollingshead, and psychiatrist Madison Presnell as medical adviser) ran psilocybin-assisted group psychotherapy for 32 volunteer prisoners at MCI-Concord, a maximum-security prison for young offenders, between about February 1961 and January 1963 [source: Wikipedia, Concord Prison Experiment; Doblin 1998]. The stated aim was to change how prisoners thought and lived so that they stayed out of prison after release, judged against the institution's recidivism base rate [source: Leary et al. 1965 abstract via Erowid; Metzner & Weil 1963 base-rate paper as cited].

Core idea of change as the authors framed it: Leary's earlier work reported that most people given psilocybin in a supportive setting said they learned something valuable about themselves, and he speculated that such experiences, inside a supportive group, could catalyse behaviour change in people with criminal records [source: Doblin 1998]. Doblin's verdict after re-examining the data is that the programme refutes the idea of a single experience as a "magic bullet": lasting change, if any, would need sustained work and post-release support [source: Doblin 1998].

What a clinician does with this knowledge today [synthesis]: (1) recognise Concord-style claims ("one experience cut reoffending in half") and know they were not supported; (2) apply its lessons when reviewing, referring to or taking part in research or programmes involving captive or dependent populations; (3) give honest, non-dosing harm-reduction and integration support to justice-involved clients who ask about psychedelics.

Legal status. United States: psilocybin is a Schedule I controlled substance under the federal Controlled Substances Act. That category means high abuse potential, no currently accepted medical use and no accepted safety under medical supervision. Possessing, making or supplying it without federal authorisation is a federal crime, and it cannot be prescribed. Research needs state and institutional approval, a DEA Schedule I registration for the specific protocol and, for human trials, an FDA investigational new drug (IND) application. Psilocybin therapies are still investigational, not FDA-approved. FDA has given breakthrough-therapy designation to COMPASS Pathways (treatment-resistant depression) and Usona Institute (major depressive disorder). Oregon (2020 ballot measure, licensing from January 2023) and Colorado run state-legal regulated programmes, and some localities have decriminalised; none of this changes federal status [source: GAO-24-106630, May 2024]. United Kingdom: psilocin and psilocybin are Class A drugs under the Misuse of Drugs Act 1971. Section 21 of the Drugs Act 2005 made all psilocin-containing ("magic") mushrooms Class A in whatever form, and the 2005 amending regulations placed such fungi in Schedule 1 of the Misuse of Drugs Regulations 2001 [source: Explanatory Memorandum to SI 2005/1653]. Schedule 1 status means psilocybin cannot be lawfully prescribed and research needs a Home Office licence. An MHRA trial approval is not a marketing approval [source: Baker McKenzie, 2023]. Elsewhere [source: Drug Science, 2025]: in Australia, authorised psychiatrists may prescribe for defined indications from 1 July 2023. Canada allows access through the Special Access Programme (2022), and Alberta regulates psychiatrist prescribing. Denmark restricts psilocybin to medical and scientific use. The law changes, so check the current position before advising. Psilocybin was not illegal in the US when Concord ran [source: Harvard Department of Psychology].

2. Framework

Historical programme structure (described, not for use) [source: Doblin 1998, summarising Leary et al. 1965; Wikipedia]:

Population and setting: 32 volunteer prisoners, mostly close to their parole dates, over about two years in one prison.
Selection: volunteers, with prisoners helping decide which other prisoners could join. No random allocation and no concurrent control group; the comparison was the institution's historical base rate (Metzner & Weil 1963).
Group format: about four prisoners and two team members per group; most groups included one prisoner who had already been through the cycle, as a peer. A group leader usually also took the drug.
Phase 1, pre-testing with feedback: personality tests (MMPI, CPI, TAT; Wikipedia also lists the Maher Sentence Completion Test) given before treatment, with results discussed with subjects.
Phase 2, non-drug preparation: group meetings to discuss test results and prepare for the drug experience.
Phase 3, day-long group drug session(s) inside the prison, followed by Phase 4, integration meetings to discuss what had happened. The whole in-prison programme ran for roughly six weeks of group meetings.
Phase 5, post-testing with the same instruments.
Phase 6, post-release follow-up: group meetings after parole approval and a non-profit (Freedom Inc.) to coordinate help. Leary later said this phase "was never fully developed"; in practice contact was irregular and individual.
Criteria for moving between phases: not reported in the sources used beyond parole timing [uncertain].
Outcome measure: return to prison compared with the base rate; personality-test change.

What was claimed and what the follow-up found [source: Doblin 1998, MAPS Bulletin]:

January 1963: Leary reported 32% recidivism against a 56% base rate. Doblin showed this compared the treated group at about 10 months after release with base-rate figures at 30 months; at a matched 10 months the base rate was about 34%, so there was no meaningful difference.
July 1964: recidivism was 59%, no different from the base rate. Leary then argued most returns were technical parole violations rather than new crimes. Doblin found many "parole violations" were in fact arrests for new crimes that later produced convictions, because only the first reason for return had been counted.
Doblin located records for 21 of 32 subjects: at 2.5 years recidivism was 71%; his estimated range for all 32 (56–88%) includes the base rate. He found no reduction at any point the base-rate data allow.
Personality tests: CPI changes on 12 of 18 scales; little MMPI change [source: Doblin 1998].
Metzner (co-investigator) accepted the errors, attributing them to a "halo effect" and carelessness rather than deliberate manipulation [source: Metzner, MAPS Bulletin 1999/2000, via MAPS page].

Modern framework for the clinician [synthesis]: there is no treatment to deliver. Use the case in three settings: (a) evaluating outcome claims; (b) ethics review of, referral to, or participation in research and programmes with captive populations (45 CFR 46 Subpart C in the US); (c) assessment, harm reduction and integration with clients, especially justice-involved clients, who ask about psychedelics. See section 4B.

3. Guiding principles

Principles the investigators stated or practised (historical; reported via Doblin 1998) [source: Doblin 1998]:

Flattened hierarchy: the team tried to avoid doctor–patient and researcher–subject roles and to share control. Practical meaning then: prisoners had a say in group decisions. Lesson now: flattening roles does not remove the power gap between free researchers and prisoners near parole [synthesis].
Solidarity through co-ingestion: a leader usually took the drug "as a demonstration of solidarity and trust" [source: Doblin 1998]. Lesson now: an impaired investigator cannot monitor safety or hold boundaries. In 1962 the Massachusetts Food and Drug Division ruled that the drugs could be given only in a physician's presence [source: Harvard Crimson, 9 July 1963]. Not acceptable practice.
Peer involvement: graduates of the cycle joined new groups. Lesson now: peer support can help, but peers are not a substitute for trained, sober staff [synthesis].
Transparency of testing: test results were fed back to subjects. Practical meaning: assessment is shared with the client rather than held over them.
Supportive set and setting: Leary wrote about set and setting (Doblin notes this in his conclusion). Practical meaning: expectations and environment shape the experience; the setting alone does not produce change.
Post-release support as the key (Leary's own later conclusion): Leary told Doblin the weak point was post-release support and later recommended an AA-style halfway house [source: Doblin 1998].

Principles a clinician should take from the case [source: 45 CFR 46 Subpart C; Doblin 1998; synthesis where marked]:

No magic bullet: a single intense experience does not by itself change entrenched behaviour; plan for sustained follow-through [source: Doblin 1998].
Voluntariness under confinement: incarceration-related pressures limit free choice; any benefit offered must not impair the person's ability to weigh risks [source: 45 CFR 46.302, 46.305].
Parole independence: participants must be told in advance, and it must be assured, that participation will not affect parole [source: 46.305].
Fair selection: selection free from arbitrary intervention by staff or prisoners; controls chosen randomly unless justified in writing [source: 46.305]. Peer-chosen entry at Concord would fail this test [synthesis].
Follow-up care must be real: if follow-up is needed, it must be provided for, allowing for different sentence lengths, and participants told [source: 46.305].
Matched comparisons and fixed definitions: compare like time windows and define outcomes before looking at data [source: Doblin 1998].
Report null results plainly: Doblin notes the 1964 null result was mentioned only once in a later paper; selective reporting misleads for decades [source: Doblin 1998].
Staff stay sober and within role: monitors who are not impaired, two-person teams and explicit touch and boundary policies are now standard in regulated programmes [synthesis; see Harm account for the 2024 MDMA-trial example].
Higher ethical bar for a contested field: Doblin argued psychedelic researchers must meet a "higher standard" to keep public and regulator trust; Metzner disagreed, saying ordinary honesty standards suffice [source: MAPS Bulletin]. Name this disagreement rather than resolve it.

4. Intervention inventory

The inventory has two parts. 4A lists every programme component the sources describe, as history (NOT FOR USE), with what went wrong. No session scripts or administration details exist here; example language is withheld for the historical components because verbatim scripts are left out for safety and none of the sources supplies therapist wording. 4B gives the procedures a clinician can actually use today; these are derived from the case and from 45 CFR 46 Subpart C and are marked [synthesis] where they are mine.

4A.1 Volunteer recruitment near parole, with peer say over entry (historical, not for use)

Purpose (as practised): enrol motivated prisoners shortly before release so outcomes could be tracked on parole [source: Doblin 1998].
Steps (as described): prisoners volunteered; existing prisoners helped decide who else joined. No random selection and no concurrent controls [source: Doblin 1998; Harvard Psychology on non-random selection in the wider project].
What went wrong: proximity to parole creates pressure and hope of favour; peer gatekeeping is the "arbitrary intervention" by prisoners that Subpart C now forbids; non-random entry made any comparison with the base rate unreliable [source: 45 CFR 46.305; synthesis].
Watch for today: recruitment timed around parole, sentencing, diversion or treatment-court decisions; selection by staff or inmate leaders.

4A.2 Pre- and post-treatment testing with feedback (historical)

Purpose: measure personality change and use results in the group [source: Doblin 1998].
Steps: MMPI, CPI and TAT (plus a sentence-completion measure per Wikipedia) before and after; results discussed with subjects. Instruments are licensed; items are not reproduced here.
What went wrong: results were reported only as group averages, so individual follow-up was impossible decades later; CPI change was presented as success though it did not translate into lower recidivism [source: Doblin 1998].
Watch for today: self-report or attitude change offered as proof of behavioural outcome.

4A.3 Non-drug preparation meetings (historical)

Purpose: discuss test results and prepare participants for the drug experience [source: Doblin 1998].
Steps: small-group meetings in the prison before the drug session. Their content is not described in the sources used [uncertain].
What went wrong: Doblin's full text describes the subjects only as volunteers, almost all close to their possible parole dates. Under Leary's principle of full disclosure they received their own test results. The text describes no consent form, no information given before enrolment and no ethics review. It does not say whether prisoners were told that taking part would or would not affect parole [source: Doblin 1998, JPD text]. What prisoners were actually told is therefore unknown [uncertain].
Contrast today: regulated programmes run preparation as part of formal informed consent, medical and psychiatric screening and safety planning [synthesis].

4A.4 Day-long group drug session with a co-ingesting leader (historical, NOT FOR USE; harm-linked)

Purpose (as stated): a shared, non-hierarchical experience; the leader's co-use was framed as solidarity and trust [source: Doblin 1998; Wikipedia].
Steps: described only in general terms: a small group of prisoners, two team members and often a peer graduate spent a day together in the prison during the drug effects. No further procedural detail is given here.
What went wrong: an investigator under the drug cannot provide safety monitoring or hold boundaries. Leary later described losing control of a group, feeling ashamed and collapsing, feeling "completely vulnerable" [source: Doblin 1998]. After the March 1962 controversy, the Massachusetts Food and Drug Division ruled that the drugs could be given only in a physician's presence [source: Harvard Crimson, 9 July 1963]. Wikipedia adds that the physician had to be sober and dates the ruling to April 1962 [source: Wikipedia, Harvard Psilocybin Project; uncertain]. Harvard faculty criticised researchers who took the drug in their own sessions, and a 2016 Crimson retrospective notes that administrators sometimes took the drugs themselves [source: Harvard Psychology; Harvard Crimson 2016].
Watch for today: any facilitator, guide or researcher who uses a substance with the client ("I journey with you").

4A.5 Post-session integration discussions (historical)

Purpose: talk through the experience and plan for the next cycle [source: Doblin 1998].
Steps: several non-drug group meetings followed the drug session(s) for discussion and integration; timing details are deliberately omitted.
Lesson: the integration element is the part that survives in modern regulated protocols; at Concord it ended at the prison gate [synthesis].

4A.6 Post-release group and Freedom Inc. (historical)

Purpose: support parolees after release [source: Doblin 1998].
Steps (planned): group meetings after parole approval and a non-profit to coordinate help. Delivered: irregular individual contact only; Leary said the phase "was never fully developed".
What went wrong: the component the investigators themselves later called essential was missing. One returned participant told Doblin he would not have gone back with post-release help [source: Doblin 1998].

4A.7 Outcome reporting against a historical base rate (historical; harm-linked)

Purpose: show reduced recidivism [source: Doblin 1998].
Steps (as done): treated-group returns at about 10 months compared with base-rate returns at 30 months; returns classified by the first reason only; later null result reframed as "technical violations".
What went wrong: see section 2 and the Harm account. Doblin calls Concord "an object lesson in what not to repeat".

4B.1 Checking an outcome claim the Concord way [synthesis, method derived from Doblin 1998]

Purpose: decide whether a programme's claim of reduced reoffending, relapse or symptoms is credible before you refer to it or repeat it.
Steps: (1) Ask what the comparison is: concurrent randomised control, matched cohort, or a historical base rate. (2) Check that follow-up windows match exactly between groups. (3) Check outcome definitions: are "technical" returns, new arrests and convictions all counted, and was the definition fixed before data were seen? (4) Count missing participants and ask how they were handled. (5) Look for later updates: were null or worse results published as prominently as the early win? (6) Ask who collected and who analysed the data, and whether they had a stake.
Example language (illustrative): "Before I recommend this, can you show me how many people were followed and for how long, compared with whom?" / to a client: "The early numbers looked good, but when someone checked the records later, the effect disappeared. Let's look at what's actually known."
Watch for / working when: you can state the comparison, window and definition in one sentence; if you cannot, treat the claim as unverified.

4B.2 Reviewing or taking part in research with incarcerated or captive people [source: 45 CFR 46 Subpart C; synthesis for steps beyond the rule]

Purpose: prevent the selection, consent and follow-up failures seen at Concord.
Steps: (1) Confirm the study fits a permitted category (46.306) and was reviewed by an IRB whose majority has no tie to the prison and which includes a prisoner or prisoner representative (46.304). (2) Check that advantages offered (better conditions, attention, hope of favour) do not impair weighing of risk (46.305). (3) Check selection is free of staff or inmate intervention and controls are random unless justified in writing. (4) Confirm participants are told in advance, in understandable language, that participation will not affect parole, and that this is assured. (5) Confirm follow-up care is planned for people released early or late. (6) Ask how adverse events are recorded and by whom independent of the investigators [synthesis]. (7) Extend the same scrutiny to other dependent groups (people under court orders, in residential treatment, students of the investigators, employees) [synthesis; the 1962 Harvard concerns included pressure on graduate students, source: Wikipedia, Harvard Psilocybin Project].
Example language (illustrative): "What exactly will prisoners be told about parole, and who guarantees it?" / "Who outside the study team decides which adverse events are reported?"
Watch for: recruitment by custody staff, incentives tied to release, no plan after release.

4B.3 Assessing a justice-involved client who asks about psychedelics [synthesis]

Purpose: respond honestly, safely and within the law, without facilitating illegal use.
Steps: (1) Explore what they have heard (often Concord-type claims) and what they hope for. (2) Give accurate information: the Concord recidivism effect was not supported on re-analysis [source: Doblin 1998]; regulated access exists only in some jurisdictions and under specific frameworks [source: Drug Science 2025]. (3) Screen generally for risks the regulated protocols exclude for (personal or family history of psychosis or bipolar disorder, serious cardiovascular disease, current interacting medicines, unstable housing or crisis) and refer medical questions to a prescriber; give no dosing or sourcing information. (4) Note legal exposure for someone on probation or parole (drug testing, conditions) and discuss it plainly. (5) If they intend to use outside a clinic, do harm-reduction planning (not alone, a sober trusted person, no driving, a plan for distress, crisis numbers) and offer integration sessions afterward. (6) Re-centre on what reduces reoffending and relapse in practice: sustained treatment, housing, work, family and peer support.
Example language (illustrative): "I can't help you get or use it, and I won't judge you for asking. I can tell you what's known and help you stay safe whatever you decide." / "The famous prison study is often quoted, but the follow-up showed people went back to prison at the usual rate."
Watch for: magical expectations, parole risk, psychosis vulnerability, pressure from a group or facilitator.

4B.4 Supporting a client harmed in a research, retreat or group-programme setting [synthesis]

Purpose: help someone who took part in a study or programme marked by coercion, boundary violations, broken follow-up promises or misrepresented results.
Steps: (1) Let them tell the story at their pace; assess for trauma symptoms, shame, self-blame, dissociation, persisting perceptual or mood changes, and suicide risk using standard risk assessment. (2) Name the power context (captivity, parole hopes, dependence on the investigator) as a real constraint on consent. (3) Separate what the experience meant to them from what the researchers claimed. (4) Explain options to report: the IRB or ethics committee, the sponsor, the national research-protection regulator (OHRP in the US), the professional licensing board, and police where a crime occurred. (5) Offer evidence-based treatment for any resulting disorder. (6) Document carefully.
Example language (illustrative): "Agreeing to take part while you were locked up and hoping for parole isn't the same as a free choice." / "You can value parts of what happened and still say other parts were wrong."
Watch for: idealisation of the investigator alternating with betrayal; re-traumatisation from retelling; legal-process stress.

4B.5 Responding to a colleague or facilitator who uses with clients [synthesis]

Purpose: act on the Concord co-ingestion red flag in current practice.
Steps: (1) Note the facts (who, when, what was said). (2) Raise it with the colleague if safe, naming client-safety and boundary concerns. (3) Consult your supervisor or ethics lead. (4) Report to the licensing board or programme sponsor as your code requires, especially if touch, sexual contact or impairment during a session is involved. (5) Support any affected client to make their own report.
Example language (illustrative): "If you're under the influence too, who is keeping the client safe?"
Watch for: framing co-use as solidarity, equality or "going deeper together".

4B.6 Building continuity after an intense experience or release [synthesis, from Leary's and Doblin's stated lesson]

Purpose: supply the post-release or post-session support Concord lacked.
Steps: before any planned change point (release, discharge, end of a programme), set up named contacts, a schedule of follow-up sessions, peer or mutual-aid groups, housing and work referrals, and a relapse or crisis plan; check in early after release.
Example language (illustrative): "Insight in here is a start. What will help you hold onto it in the first month outside?"
Working when: the client keeps appointments after the change point and has at least one support they actually use.

5. Handling problems

Problems the historical sources record [source: Doblin 1998; MAPS Bulletin]:

Loss of control of a group: Leary admitted losing control of a group and collapsing while under the drug. Response today: staff in any session with altered states stay sober, work in pairs and have a written escalation and medical plan; medical monitoring belongs to medical staff [synthesis].
Non-response after release (recidivism at base rate): the investigators' own explanation was absent post-release support. Response: build continuity before release (4B.6) and do not treat in-custody insight as an outcome [source: Doblin 1998; synthesis].
Missing records and lost participants: only 21 of 32 files could be found decades later. Response: plan long-term follow-up and data retention at the start; report attrition [synthesis].
Investigator bias: Metzner attributed the errors to a "halo effect" [source: MAPS Bulletin]. Response: independent outcome assessment, pre-specified definitions, external audit [synthesis].

Problems the sources do not address [all synthesis]:

Client idealises Leary or the Concord claim: don't argue; share the follow-up findings, ask what the claim means to them (hope of change, wish for a shortcut) and work with that hope.
Client in custody or on parole wants to join a psychedelic study: go through 4B.2 with them: what they will be told about parole, who selects, what follow-up exists after release, whether they can withdraw without penalty. Encourage them to ask the study team in writing.
Client intends to use with an underground guide who "journeys with" clients: name the co-use red flag, do harm-reduction planning without dosing or sourcing, and offer integration; do not facilitate.
Rupture when you challenge a cherished belief or a facilitator the client trusts: acknowledge their experience is real to them, separate it from the facilitator's claims, repair by returning to the client's own goals.
Client distressed after taking part in a study (anxiety, intrusive memories, perceptual changes, shame): standard assessment including suicide risk; stabilisation; evidence-based trauma or anxiety treatment; medical referral for persisting perceptual symptoms; support to report.
Overwhelm on learning a programme they credited was flawed: pace the information; validate that their own change may be real even if the study's claim was not.
Non-adherence to follow-up after release: reduce barriers (outreach, flexible scheduling, linking with probation-approved services) and keep the door open.
You are asked to help design or promote a prison psychedelic project: insist on Subpart C safeguards, independent adverse-event reporting, sober two-person teams, a touch and boundary policy, pre-registration and a funded post-release component.

Harm account

What was actually done

Who, where, when: Harvard's Center for Research in Personality team led by Timothy Leary, at MCI-Concord, Massachusetts, about February 1961 to January 1963, with 32 prisoners [source: Wikipedia; Doblin 1998].
Selection and consent: volunteers, almost all close to their possible parole dates. Other prisoners helped decide who joined. Doblin's full text records no consent form, no information given before enrolment and no ethics review. It does not say whether prisoners were told how taking part related to parole [source: Doblin 1998, JPD text]. Leary himself believed that the extra supervision from post-release contact produced more technical parole violations [source: Doblin 1998]. What prisoners were told, and whether parole boards took part in the decisions, is still unknown [uncertain].
Sessions: small mixed groups (prisoners, two team members, often a graduate peer); testing with feedback; non-drug preparation; day-long group drug sessions in the prison in which a group leader usually also took the drug; integration meetings; post-testing [source: Doblin 1998].
After release: the promised group meetings and support organisation were never properly delivered; contact was irregular [source: Doblin 1998].
Reporting: the January 1963 claim of 32% against 56% rested on mismatched follow-up windows. The July 1964 null result (59%) was reframed as mostly technical violations, though many such returns were new-crime arrests. The null result was later mentioned only once in print [source: Doblin 1998]. Leary et al. published the programme in 1965 as "a new behavior change program" [source: Erowid abstract].
The wider Harvard controversy: the Harvard Psilocybin Project (from 1960) drew faculty concern over subject safety, non-random selection, poor controls and researchers taking the drug in their own sessions [source: Harvard Psychology]. Wikipedia dates the first Harvard Crimson report of these concerns to 20 February 1962 [source: Wikipedia, Harvard Psilocybin Project] [uncertain: secondary, unverified]. In March 1962, at a closed meeting of the Center for Research in Personality, members charged that the experiments were unscientific and irresponsible. Leary and Alpert pledged not to give the drugs to undergraduates, and the Massachusetts Food and Drug Division ruled that the drugs could be given only in a physician's presence [source: Harvard Crimson, 9 July 1963]. Psychology professor Herbert Kelman circulated a letter and called a meeting to stress that taking part had to be voluntary. Allan Cohen, a graduate student who took part, later recalled that students could choose between the final exam and the psilocybin trials [source: Harvard Crimson, 2016 retrospective]. On 26 November 1962, Dean Monro and Dr Dana Farnsworth of University Health Services warned undergraduates in a letter to the Crimson that the drugs were a serious mental-health hazard. On 10 December, Leary and Alpert replied that the University's position was scientifically reckless and inaccurate [source: Harvard Crimson, 9 July 1963]. On 27 May 1963, President Nathan Pusey announced that the Harvard Corporation had terminated Richard Alpert's appointment for giving a hallucinogenic drug to an undergraduate against his agreement with the University. It was the first dismissal in Pusey's ten years as president. The Corporation also relieved Leary of his teaching duties and suspended his salary for leaving the University without permission. The Crimson's 28 May editorial backed the dismissal as a response to "willful repudiation" of scientific safeguards [source: Harvard Crimson, 9 July 1963]. A 2016 Crimson retrospective says instead that Leary's contract was not renewed after he failed to meet a semester's teaching duties [source: Harvard Crimson 2016]; this guide follows the contemporaneous account. Wikipedia reports that two students were hospitalised in 1961 and dates the undergraduate agreement to 1961; neither Crimson article opened confirms the hospitalisations [source: Wikipedia; uncertain]. Harvard's psychology department notes that the researchers lost credibility for weak rigour and for not following research guidelines [source: Harvard Psychology]. The Harvard sources describe the Concord study as the more academically sound part of the work [source: Wikipedia].

How it caused harm

No documented physical injuries or deaths from Concord appear in the sources used; adverse reactions were not systematically reported, which is itself a failure [source: Doblin 1998]. The two participants Doblin interviewed (with an investigator present) said it helped and reported no lasting harm; two of 32 cannot establish safety [source: Doblin 1998; synthesis].
Compromised voluntariness: prisoners near parole, selected in part by peers, with no recorded consent process, in a closed institution; these are the exact risks Subpart C later addressed [source: 45 CFR 46 Subpart C; synthesis].
Impaired monitors: a leader under the drug cannot protect participants; Leary's own account of collapse in a group shows the risk was realised at least once [source: Doblin 1998].
Broken promise of support: participants were offered post-release help that did not materialise; recidivism reached 71% at 2.5 years in the traceable sample, and one returned participant tied his return to the lack of help [source: Doblin 1998].
Misinformation for decades: the 1963 claim was widely repeated; Doblin's re-analysis came 34 years later. Overclaiming can raise false hopes in prisoners and families, divert policy and funding, and invite backlash [source: Doblin 1998; synthesis].
Damage to the field: the controversy ended Harvard's research; Doblin argued that such lapses cost the trust of regulators and the public [source: Harvard Psychology; Doblin 1998].
Most at risk: young prisoners close to parole, those without family or post-release support, students dependent on the investigators [synthesis from the sources above].

Recognising it today

Features and phrases that echo Concord [synthesis unless cited]:

Claims that one or a few intense experiences reduce reoffending, relapse or trauma permanently, compared against a historical or national base rate.
Follow-up windows that differ between groups, outcomes redefined after the data ("only technical violations"), missing participants not reported, null updates buried.
Recruitment in prisons, treatment courts, residential programmes or among one's own students or staff, especially with hints of favourable treatment.
Facilitators who use with participants ("solidarity", "we go together", "no hierarchy here").
Strong investigator allegiance with no independent outcome assessment.
Promised aftercare that is vague or unfunded.

A recent example of related concerns: in 2024 the Institute for Clinical and Economic Review (ICER) questioned MDMA-assisted therapy trials for functional unblinding, "very strong prior beliefs" among therapists and investigators, and reports that participants were encouraged to report favourably; a citizen petition alleged under-reporting of adverse events (disputed by MAPS and the trialists). MAPS had earlier found that two therapists at a Canadian Phase 2 site "substantially deviated" from the manual through inappropriate physical contact with a participant (Meaghan Buisson) under the drug, barred them and notified Canadian and US authorities [source: NPR via VPM, 13 May 2024]. On 4 June 2024 an FDA advisory committee voted 9–2 that effectiveness was not shown and 10–1 that benefits did not outweigh risks [source: KAXE/NPR, 4 June 2024]. These are allegations and judgements about a different programme; they are cited as parallels, not equivalents.

How legitimate versions differ [source: 45 CFR 46 Subpart C; synthesis]: independent ethics review with prisoner representation; random or justified selection; written assurance on parole; a documented consent process; sober, trained two-person teams with explicit touch and boundary policies; medical monitoring by medical staff; independent adverse-event reporting; pre-registered outcomes with matched follow-up; funded aftercare; publication of null results.

What to do instead and how to help those harmed

Instead, for reducing reoffending: structured, sustained rehabilitation that continues into the community. Risk–need–responsivity (RNR) is described as the leading paradigm for interventions targeting recidivism. It means assessing criminogenic risk, targeting criminogenic needs such as criminal thinking, and fitting delivery to the person, for example with Motivational Interviewing. Named structured cognitive-behavioural programmes include Thinking for a Change (T4C), Moral Reconation Therapy (MRT), Reasoning & Rehabilitation (R&R) and Interactive Journaling. DBT and Schema-Focused Therapy are named for people with mental illness [source: NIC library, Reducing Criminal Recidivism for Justice-Involved Persons with Mental Illness]. A methodological quality review concluded that MRT and R&R "appear to be successful approaches to reducing recidivism" [source: Allen, MacKenzie & Hickman 2001, RAND abstract]. Add substance-use treatment, housing, employment support and mutual-aid groups as needed, and deliver named programmes only with the training their developers require [synthesis]. Doblin and Leary both pointed to post-release support (Leary proposed an AA-style halfway house) [source: Doblin 1998].
Instead, for psychedelic research with prisoners: Doblin proposed a new controlled study combining group therapy with a post-release programme [source: Doblin 1998]; any such study should meet all Subpart C findings [source: 45 CFR 46.305–306].
Helping those harmed: use 4B.4: trauma-informed assessment including suicide risk, validation of constrained consent, separating the experience from the claims, evidence-based treatment for resulting disorders, medical referral for persisting perceptual symptoms [synthesis].
Duties: report research misconduct or participant harm to the IRB or ethics committee and the sponsor. In the US, OHRP oversees HHS-funded human-subjects protections. Report boundary or sexual misconduct by licensed professionals to the licensing board and follow mandatory-reporting law. Support clients' own complaints and keep records [synthesis; the MAPS example shows notification of national health authorities, source: NPR via VPM].

Sources

Pages consulted:

Pages consulted in the October 2026 verification:

Works cited:

Leary, T., Metzner, R., Presnell, M., Weil, G., Schwitzgebel, R., & Kinne, S. (1965). A new behavior change program using psilocybin. Psychotherapy: Theory, Research & Practice, 2(2), 61–72. DOI 10.1037/h0088612. Known here through its abstract and Doblin's account [uncertain].
Metzner, R., & Weil, G. (1963). Predicting recidivism: base-rates for Massachusetts Correctional Institution Concord. Journal of Criminal Law and Criminology, 54(3), 307–316. DOI 10.2307/1140984; journal archive copy (Northwestern Scholarly Commons). [uncertain on content]
Doblin, R. (1998). Dr. Leary's Concord Prison Experiment: a 34-year follow-up study. Journal of Psychoactive Drugs, 30(4), 419–426. Full text: MAPS bibliography copy and MAPS Bulletin reprint.
Metzner, R. (1999/2000). Reflections on the Concord Prison Experiment and the follow-up study. MAPS Bulletin, 9(4), 19–20. MAPS Bulletin page above.
Allen, L. C., MacKenzie, D. L., & Hickman, L. J. (2001). The effectiveness of cognitive behavioral treatment for adult offenders: a methodological, quality-based review. International Journal of Offender Therapy and Comparative Criminology, 45(4), 498–514. RAND abstract page.

← Previous · VII-038Naranjo's MDA, harmaline and ibogaine therapy (historical)Next · VII-040 →Spring Grove research (historical)