Nine groups of rats self-administered intravenously-delivered cocaine (0.1, 0.2, or 0.4 mg/kg) during 24-h sessions contingent upon lever-press responses under a fixed-ratio (FR) 4 schedule. Three other groups of rats responded on tongue-operated drinking devices for deliveries (0.01 ml) of a solution of glucose and saccharin (G + S). There were an additional three groups that initially self-administered cocaine (0.2 mg/kg), and later saline replaced cocaine and extinction behavior was allowed to stabilize. All 15 groups of rats were injected twice daily for 5 days with one of three doses of buprenorphine (0.1, 0.2 or 0.4 mg/kg). Buprenorphine decreased cocaine self-administration, but the effect of the highest dose was only slightly greater than that of the lowest dose tested. Cocaine infusions were reduced on the first day of treatment, but they increased over the next 4 days of buprenorphine injections. Buprenorphine decreased G + S intake during the last 2 or 3 days of injections. When buprenorphine treatment was terminated, G + S intake decreased even further. These lower rates of intake persisted for at least 5 days, and they returned to baseline by 2 weeks. Saline self-administration was decreased by buprenorphine in all saline extinction groups. Food intake was not altered by buprenorphine in the groups self-administering IV cocaine or saline; however, food intake was reduced in the G + S groups. Water intake increased during buprenorphine treatment in some of the cocaine groups but not in the G + S groups.(ABSTRACT TRUNCATED AT 250 WORDS)