Thirty-nine methadone maintenance patients were included in a 9-day, double blind, randomized, inpatient detoxification trial. Methadone was tapered to 10 mg/day and then patients were assigned to one of these 3 protocols: clonidine (0.3-0.9 mg/day), lefetamine (60-240 mg/day), buprenorphine (0.15-0.9 mg/day). Buprenorphine treatment was significantly superior to clonidine and to lefetamine (F = 3.96 df = 2, 29 P < 0.05) in controlling objective, subjective and psychological withdrawal symptomatology. Clonidine was more effective than lefetamine in suppressing withdrawal in the first 3 days of treatment (day 3: F = 4.10 df = 2, 30 P < 0.05), and this trend was apparent on the objective and psychological items. In addition to evaluations of the efficacy of the single drugs used, the study showed that tapering methadone to low doses before entering the pharmacologically assisted discontinuation phase was clinically acceptable in detoxification from long-term methadone treatment.