BACKGROUND It has been proposed that relapse vulnerability in previously dependent individuals results from augmentation of drug-induced reinforcement due to repeated associations between the interoceptive properties of the drug and reduction of acute withdrawal distress. METHODS To test this hypothesis, male Sprague-Dawley rats self-administered 0.05 mg/kg/inf heroin on continuous reinforcement (CR) and progressive ratio (PR) schedules. During this period, they also received injections of vehicle or escalating doses of heroin. Following tests of naloxone-precipitated withdrawal, as well as a drug-free period (4 days), and extinction (9 sessions), they were pre-treated with vehicle or yohimbine (0.5mg/kg, IV) and tested for resumption of heroin self-administration (0.05 mg/kg/inf) on CR and PR schedules, or tested for reinstatement in extinction conditions. RESULTS Increased self-administration on the CR schedule was observed in the heroin-injected rats, but no group differences were observed on the PR schedule, in spite of greater signs of withdrawal precipitated by naloxone in the heroin-injected rats. More importantly, there were no group differences in resumption of heroin self-administration, and this was not altered by yohimbine. CONCLUSIONS These results suggest that relapse vulnerability cannot be uniquely ascribed to enhanced reinforcing action of drugs; contextual and other conditioning factors must play a role in modulating resumption of drug intake after abstinence.