Tihar Vipassana began in November 1993, when 96 inmates and 23 jail staff sat a ten-day silent course inside Tihar Jail in Delhi under Inspector General of Prisons Kiran Bedi. A second course in January 1994 took 300 inmates, and in April 1994 S. N. Goenka led a course for more than 1,000 prisoners, the largest Vipassana course held up to then, after which a permanent centre, Dhamma Tihar, opened inside the jail. The story is real; the evidence that it reformed prisoners is thinner than the legend, though controlled studies elsewhere do show modest benefits.
- November 1993: the first Tihar course, with 96 inmates and 23 jail staff.
- April 1994: more than 1,000 inmates sat one course, then the largest Vipassana course held.
- Prison Vipassana in India began earlier, at Jaipur Central Jail in 1975.
- The best controlled study (Seattle, 2006) found less drug and alcohol use but no difference in re-arrest.
- Across prison yoga and meditation studies, wellbeing rose by d = 0.46: real, moderate, not transformation.
The planning brief for this post said April 1993 for the big course. The records say otherwise:
- Kiran Bedi only took charge of Delhi's prisons in May 1993;
- the first Tihar course was in November 1993;
- the thousand-person course was in April 1994.
This post sets out that timeline from the Vipassana Research Institute's own history, adds what independent sources record about Tihar at the time, and then weighs the claims against the research on meditation in prisons.
What was Tihar like when the Vipassana courses began?
Tihar was massively overcrowded and run largely as a holding pen. When Bedi was posted as Inspector General of Delhi Prisons in May 1993, the complex held roughly 8,000 to 9,500 prisoners against a built capacity of about 2,500, and around 90% were undertrials, people not yet convicted of anything.
Vipassana was one reform among many
Bedi's reforms were broad. She:
- separated hardened offenders;
- brought in education through IGNOU and the National Open School;
- set up legal aid cells for undertrials;
- banned smoking, and opened a de-addiction centre and workshops;
- created prisoner panchayats and petition boxes.
That context matters for evaluation: any change in behaviour at Tihar in 1993 to 1995 had many possible causes.

How did the Tihar Vipassana courses happen?
Vipassana had been tried in Indian prisons long before Tihar. The first course for prisoners was held at Jaipur Central Jail in October 1975, arranged by Ram Singh, then Rajasthan's Home Secretary, with a second there in 1977.
Ram Singh had sat a Vipassana course himself, and as Rajasthan's Home Secretary he proposed trying it on prisoners. S. N. Goenka came to Jaipur Central Jail to teach it. The first obstacle was the rule book: outsiders were not allowed to stay inside a jail. Goenka's answer, as the Vipassana Research Institute tells it, was to ask to be sentenced to prison. Administrative orders settled it instead, and he lived in a makeshift room in the jail dispensary.
Four of the prisoners were brought to the course in ankle locks and handcuffs. Goenka objected, the chains came off, and an armed guard was posted at a strategic point instead. Two condemned prisoners could not join, so the instructions were relayed to their cells by loudspeaker. The institute records 114 inmates on that first course.
Vipassana Research Institute: Vipassana in Prisons, History and SpreadAfter a gap of nearly 15 years, seven more courses ran in jails in Rajasthan and Gujarat in 1990 and 1991. Tihar was the step from pilot to scale.
See the data as a table
| Inmates (people) | Jail staff (people) | |
|---|---|---|
| Nov 1993 | 96 | 23 |
| Jan 1994 | 300 | |
| Apr 1994 | 1000 |
Source: Vipassana Research Institute: Vipassana in Prisons, History and Spread · April 1994 is recorded as over 1,000 inmates. Staff numbers are given only for the first course.
| Date | Course | Participants | Teachers |
|---|---|---|---|
| Oct 1975 | Jaipur Central Jail, first prison course in India | Inmates (VRI gives 114) | Not stated |
| 1990–91 | Seven courses in Rajasthan and Gujarat jails | Inmates | Assistant teachers |
| May 1993 | Kiran Bedi becomes IG, Delhi Prisons | ||
| Nov 1993 | First Tihar course | 96 inmates, 23 staff | Assistant teachers |
| Jan 1994 | Second Tihar course | 300 inmates | Six assistant teachers |
| Apr 1994 | Mass course, then permanent centre Dhamma Tihar | Over 1,000 inmates | Goenka and ten assistant teachers |
| 3 May 1995 | Bedi removed as head of prisons | ||
| 1997 | 75 courses held in Indian jails that year |
How a prison course runs
Prison courses follow the standard ten-day format, as the later Seattle study describes it:
- Separate housing. Participants are housed apart from other inmates.
- Silence. They keep silence except for questions to the teacher.
- Breath first. The first days are spent observing the breath.
- Sensations from day four. They then move to systematic observation of body sensations without reacting to them.
- No fee. Courses are offered free of charge.
Staff went first: Bedi's officers sat a course before the large inmate courses, a pattern later copied in Sri Lanka, Ireland, Canada and Mexico.
What did the 1997 documentary show?
Doing Time, Doing Vipassana, a 52-minute film by Ayelet Menahemi and Eilona Ariel for Karuna Films, documented the Tihar programme and Bedi's role. It won the Golden Spire at the San Francisco International Film Festival in 1998 and an award from the US National Council on Crime and Delinquency the same year.
Its biggest effect was abroad. The film helped inspire Vipassana courses at the North Rehabilitation Facility, a minimum-security jail in Seattle, where the first North American prison course ran in 1997 with 11 inmates.
That Seattle programme became the subject of the best-designed study of prison Vipassana so far, which is how a Delhi jail ended up shaping American addiction research. Bedi received the Ramon Magsaysay Award for Government Service in 1994.

Did Vipassana actually reform prisoners at Tihar?
The honest answer is that Tihar itself was never rigorously evaluated. The claims come mainly from the Vipassana organisation's reports and a few Indian studies it cites.
- The AIIMS study (1997). The Vipassana Research Institute reports that a study of Tihar inmates by the All India Institute of Medical Sciences found less anger, tension, hostility, revenge-seeking and helplessness, fewer drug-addiction and psychiatric symptoms, and more willingness to cooperate with authorities. It is cited in later peer-reviewed work as Chandiramani, Verma and Dhar. I could not find a free full text to check its design.
- Rajasthan and Gujarat (1990s). Studies by the Gujarat education department and the University of Rajasthan are described as finding better attitudes, relations with staff and self-discipline. Again, no controlled design is reported.
- Selection. Inmates chose to attend, and authorities chose who could. People willing to sit ten days in silence are not a random sample of prisoners.
- Everything else changing. Tihar's education, legal aid and de-addiction reforms ran at the same time.
What does controlled research on prison meditation show?
It shows real but modest effects, mostly on wellbeing and substance use, and little proof on reoffending. Two studies are the anchor.
| Study | Design | Sample | Main result | Limit |
|---|---|---|---|---|
| Bowen et al. 2006, Seattle | Vipassana course vs treatment as usual | 305 at baseline; 173 after course (57 vs 116); 78 at 6 months | Less alcohol, marijuana and crack cocaine use 3 months after release; fewer psychiatric symptoms | Not randomised; self-report; no recidivism difference |
| Auty, Cope and Liebling 2017, Cambridge | Systematic review and two meta-analyses of yoga and meditation in prison | 24 studies reviewed, 13 meta-analysed | Wellbeing d = 0.46; behavioural functioning d = 0.30 | Many small, low-quality studies; programmes vary widely |
| Pandey and Aryal 2026, Nepal | Pre, post and follow-up surveys | Inmates on courses at Nakkhu Jail, 2022–24 | Lower anger, stress and anxiety; gains held at follow-up | No control group |
Seattle: the sobering numbers
The Seattle numbers are the most sobering. Six months after release, Vipassana participants averaged 0.61 bookings against 0.67 for the comparison group, a difference that was not significant, with 72% of all participants not re-arrested at all.
The authors note the base rate may have been too low to detect a difference. In other words: the course helped people drink and use less, and made them feel better, but nobody has yet shown it keeps them out of jail.
The course helped people drink and use less, and feel better. Nobody has yet shown it keeps them out of jail.Reading Bowen et al. 2006
Cambridge: the pooled picture
An effect size of d = 0.46 on wellbeing is a real, moderate effect. It is also a long way from transformation. The Cambridge authors found that longer, less intensive programmes did slightly better on behaviour than short intensive ones, which counts against the idea that a single ten-day course is enough.
Why did Tihar become the story everyone tells?
Scale, setting and a film. A thousand prisoners in silence inside one of India's most crowded jails is a powerful image, and the documentary carried it to audiences who would never read a prison department report. The programme also fitted Bedi's public image as a reforming police officer, and Goenka's movement had good reason to publicise it.
None of that makes it false. The courses happened, Dhamma Tihar opened, and the model spread to jails in Taiwan (1996), the United States (1997), the United Kingdom (1998), Nepal (1998), New Zealand (1999) and many countries since.
What the story lacks is the measurement that would tell us how much it changed lives.
If you are weighing a course for yourself, the evidence for ordinary practitioners is in whether a 10-day Vipassana course is worth it; for the broader claims, see mindfulness: the hype and the evidence; and for how meditation became a lab subject at all, the history of meditation research.
A related story shows the same gap between headline and data from the lab side: the monk brain study and its gamma waves. For more on the mind and its practices, the Manthan world collects them.

Nine ten-day Vipassana courses at the North Rehabilitation Facility in Seattle over 15 months, compared with inmates receiving treatment as usual. Of 305 who completed baseline (63 Vipassana, 242 comparison), 173 completed the post-course assessment and 78 the six-month follow-up. Three months after release the Vipassana group reported significantly less alcohol, marijuana and crack cocaine use, fewer alcohol problems and psychiatric symptoms, and more optimism. Re-arrest did not differ (0.61 vs 0.67 bookings). Limit: participants chose their group, measures were self-reported, and follow-up attrition was high.
Read the paper (free)Cambridge criminologists searched the literature to December 2014, including Vipassana studies, and pooled 13 of 24 eligible studies. Prisoners who completed yoga or meditation programmes showed a small-to-moderate rise in psychological wellbeing (d = 0.46, 95% CI 0.39 to 0.54) and a small improvement in behavioural functioning (d = 0.30, 95% CI 0.20 to 0.40). Longer, less intensive programmes did slightly better on behaviour. Limit: the underlying studies were heterogeneous and often small, and few were randomised.
Read the paper (free)What's new
- Bengaluru Central Prison runs a 10-day Vipassana camp for about 25 convicted inmates with good conduct records Country and Politics, May 2026
- Nepal study: 43 prison Vipassana courses and 1,420 participants by December 2025, with gains held at follow-up Lumbini Prabha, Apr 2026
Questions people ask
When was the Vipassana course in Tihar Jail?
The first course was in November 1993 with 96 inmates and 23 staff, the second in January 1994 with 300 inmates, and the largest in April 1994 with more than 1,000 inmates, led by S. N. Goenka. A permanent centre, Dhamma Tihar, followed.
Does Vipassana reduce reoffending?
Not proven. The best controlled study, in a Seattle jail, found less drug and alcohol use after release but no significant difference in re-arrests at six months. Meta-analyses show modest gains in wellbeing and behaviour inside prison.
Is a 10-day Vipassana course hard?
It is demanding: up to 11 hours of meditation a day, in silence, with a fixed schedule. In prison courses participants are also housed separately and cut off from outside contact for the ten days.
What was Kiran Bedi's role?
As Inspector General of Delhi Prisons from May 1993 to May 1995, she arranged the Tihar courses, had staff sit a course first, and pushed for the large inmate course. She received the Ramon Magsaysay Award in 1994.
Sources
- Vipassana Research Institute: Vipassana in Prisons, History and Spread
- Bowen S et al.: Mindfulness meditation and substance use in an incarcerated population, Psychology of Addictive Behaviors 20(3), 343–347 (2006)
- Auty KM, Cope A, Liebling A: A systematic review and meta-analysis of yoga and mindfulness meditation in prison (2017)
- Pandey MP, Aryal DR: Application of Vipassana meditation for prisoners' behavior reform: a case of Nepal, Lumbini Prabha 11 (2026)
- Ministry of Health and Family Welfare: Tele-MANAS
- Wikipedia: Kiran Bedi, Delhi Prisons 1993–95 (accessed 2026)
- Wikipedia: Doing Time, Doing Vipassana (accessed 2026)
- Country and Politics: Bengaluru Central Prison Vipassana camp (2026)
