The optimum dose of buprenorphine for acute inpatient heroin detoxification has not been determined. This randomized, double-blind, double-dummy, pilot study compares two buprenorphine sublingual tablet dosing schedules to oral clonidine. Heroin users (N = 30) who met DSM-IV criteria for opioid dependence and achieved a Clinical Opiate Withdrawal Scale (COWS) score of 13 (moderate withdrawal), were randomized to receive higher dose buprenorphine (HD, 8-8-8-4-2 mg/day on days 1-5), lower dose buprenorphine (LD, 2-4-8-4-2 mg/day on days 1-5), or clonidine (C, 0.2-0.3-0.3-0.2-0.1 mg QID on days 1-5). COWS scores were obtained QID. Twenty-four hours after randomization, the percentages of subjects who achieved suppression of withdrawal, as defined by four consecutive COWS scores <12, were: C = 11%, LD = 40%, and HD = 60%. Generalized estimating equation regression models, controlling for baseline COWS and time, indicated that COWS scores over the course of 5 days were lower in both LD and HD compared to C (chi(2)(2) = 13.28, P = 0.001). Similar analyses examining scores over time on the Adjective Rating Scale for Withdrawal (ARSW) and on a Visual Analog Scale of Opiate Craving (VAS) indicated an overall treatment effect on the VAS accounted for by a significant difference between HD and C, but no overall treatment effect on the ARSW. There were no discontinuations due to treatment-related adverse events. Both HD and LD regimens are safe and efficacious treatment for opioid detoxification, but HD demonstrated superiority to C on a greater number of measures.