The partial opiate-receptor agonist buprenorphine has been suggested for treatment of heroin dependence, but there are few long-term and placebo-controlled studies of its effectiveness. We aimed to assess the 1-year efficacy of buprenorphine in combination with intensive psychosocial therapy for treatment of heroin dependence. 40 individuals aged older than 20 years, who met DSM-IV criteria for opiate dependence for at least 1 year, but did not fulfil Swedish legal criteria for methadone maintenance treatment were randomly allocated either to daily buprenorphine (fixed dose 16 mg sublingually for 12 months; supervised daily administration for a least 6 months, possible take-home doses thereafter) or a tapered 6 day regimen of buprenorphine, thereafter followed by placebo. All patients participated in cognitive-behavioural group therapy to prevent relapse, received weekly individual counselling sessions, and submitted thrice weekly supervised urine samples for analysis to detect illicit drug use. Our primary endpoint was 1-year retention in treatment and analysis was by intention to treat. 1-year retention in treatment was 75% and 0% in the buprenorphine and placebo groups, respectively (p=0.0001; risk ratio 58.7 [95% CI 7.4-467.4]). Urine screens were about 75% negative for illicit opiates, central stimulants, cannabinoids, and benzodiazepines in the patients remaining in treatment. The combination of buprenorphine and intensive psychosocial treatment is safe and highly efficacious, and should be added to the treatment options available for individuals who are dependent on heroin.