Stress plays an important role in substance abuse problems. For example, in studies with rodents stress induces reinstatement of opioid and cocaine self-administration. In addition, attenuation of the stress response by pharmacological adrenalectomy using ketoconazole, a cortisol synthesis inhibitor, reduces cocaine self-administration in rodents. In contrast, studies in primates and humans have produced conflicting results using cortisol synthesis inhibitors for attenuating cocaine-related behaviors and subjective effects. To explore the treatment implications of these findings, ketoconazole's (600-900 mg daily) ability to reduce heroin and cocaine use was compared with placebo in 39 methadone maintained patients with a history of cocaine abuse or dependence during a 12-week double blind trial. Contrary to the predicted effects, both heroin and cocaine use increased after patients were stabilized on methadone and ketoconazole. Depressive and withdrawal symptoms improved no more with ketoconazole than with placebo treatment, and side effects were greater on ketoconazole than placebo. As reported before with methadone treatment, morning cortisol levels were significantly lower than normal values throughout the clinical trial, but were not lower with ketoconazole than placebo treatment. Thus, in agreement with the negative results from acute dosing studies in primates and humans, chronic ketoconazole treatment does not appear to reduce cocaine or opioid use in humans maintained on methadone.