Heroin self-administration behavior under a progressive ratio (PR) schedule of reinforcement was evaluated in rats. The schedule was designed to restrict drug intake, minimize opiate dependency, and quantify the number of responses emitted (final response ratio) in order to receive a limited number of heroin infusions. Final ratios were found to be stable and did not increase with chronic (31 days) PR reinforcement. The ability of the PR schedule to detect changes in heroin reinforcement was demonstrated by evaluating the effect of naltrexone pretreatment and unit dose alteration on final ratios. Naltrexone (0.4 mg/kg) reduced final ratios and an inverted U dose-response relationship was established for the unit heroin doses 12.5-100 micrograms/injection. Maximal final ratios occurred with 50 micrograms/injection heroin reinforcement. This PR schedule may provide a useful method for evaluating the effects of pharmacological manipulations or lesions on opiate reinforcement.