Several sources indicate that intravenously administered buprenorphine may have significant abuse liability in humans. The present study evaluated the reinforcing effects of intravenously administered buprenorphine (0, 2, and 8 mg) in detoxified heroin-dependent participants during a 7.5-week inpatient study. Participants (n ϭ 6) were detoxified from heroin over a 1.5-week period immediately after admission. Testing subsequently occurred in three 2-week blocks. During the first week of each 2-week block, the reinforcing effects of buprenorphine were evaluated. Participants first received a dose of buprenorphine and $20 and then were given either the opportunity to selfadminister the dose or $20 during choice sessions. During the second week of each 2-week block, the direct effects of heroin were measured to evaluate potential long-lasting antagonist effects of buprenorphine. Progressive ratio break-point values were significantly higher after 2 and 8 mg of buprenorphine compared with placebo. Correspondingly, several positive subjective ratings increased after administration of active buprenorphine relative to placebo. Although there were few differences in peak effects produced by 2 versus 8 mg of buprenorphine, the higher buprenorphine dose generally produced longer-lasting effects. Heroin also produced dose-related increases in several subjective effects. Peak ratings produced by heroin were generally higher than peak ratings produced by buprenorphine. There was little evidence of residual antagonism produced by buprenorphine. These results demonstrate that buprenorphine served as a reinforcer under these conditions, and that it may have abuse liability in nonopioid-dependent individuals who abuse heroin.Buprenorphine, a -opioid receptor antagonist and partial -opioid receptor agonist, is currently approved by the Food and Drug Administration for treating pain. It is also under investigation as a maintenance medication for the treatment of opioid dependence, for which it has demonstrated effectiveness (for review, see Bickel and Amass, 1995). Studies have shown that maintenance on approximately 8 mg of sublingual liquid buprenorphine is as effective as 50 to 60 mg of oral methadone in reducing illicit opioid use (Johnson et al., 1992;Strain et al., 1994; however, see Ling et al., 1996). One purported advantage of buprenorphine compared with methadone is that the abuse liability of buprenorphine is thought to be lower. However, several preclinical studies have demonstrated that buprenorphine is self-administered above placebo levels in both drug-experienced and -naive nonhuman primates (for review, see Negus and Woods, 1995), and some studies demonstrated that the magnitude of buprenorphine self-administration was similar to that of full -agonists. For example, the maximal rates of responding maintained by buprenorphine and the full -agonists alfentanil and heroin were equivalent in rhesus monkeys selfadministering these drugs intravenously (Winger et al., 1992; but see Winger and Woods, 2001). Another...