Effective treatment of opioid dependence with methadone and of tobacco dependence with nicotine illustrates the potential usefulness of agonist medications for drug abuse treatment. The monoamine-releaser d-amphetamine is one drug under consideration as an agonist pharmacotherapy for cocaine dependence. To assess the concordance between preclinical evaluations and ongoing clinical trials, the present study examined the effects of chronic treatment with saline or d-amphetamine on cocaine- and food-maintained responding in rhesus monkeys. Cocaine injections and food pellets were available under a second-order schedule during alternating daily sessions of cocaine and food availability. d-Amphetamine (0.01-0.1 mg/kg per h i.v. for 7 consecutive days) dose-dependently decreased self-administration of a unit dose of cocaine (0.01 mg/kg per injection) at the peak of the cocaine self-administration dose-effect curve. d-Amphetamine (0.032-0.1 mg/kg per h for 7 days) also decreased self-administration of a broad range of cocaine doses (0.0032-0.1 mg/kg per injection) and produced rightward and downward shifts in the cocaine dose-effect curve. Food-maintained responding was usually decreased less than cocaine self-administration, and few signs of toxicity were noted. To evaluate the effects of a longer treatment regimen, d-amphetamine (0.1 mg/kg per h) was administered for 28 consecutive days. d-Amphetamine nearly eliminated self-administration of cocaine (0.01 mg/kg per injection) throughout this treatment, whereas food-maintained responding returned to baseline levels after approximately 9 days. These preclinical findings are concordant with recent clinical studies and suggest that chronic d-amphetamine may selectively decrease cocaine-taking behavior in rhesus monkeys, possibly by producing a selective decrease in the reinforcing effects of cocaine.