The present study describes a procedure that permits rapid assessment of environmental and pharmacological factors that may influence the choice between cocaine and food in rhesus monkeys. Daily 2 h sessions were divided into five components. During each component, monkeys (N=4) chose between i.v. cocaine (0-0.1 mg/kg/injection) and food (0, 1, or 3 food pellets). Up to 10 reinforcers were available during each component, and different discriminative stimuli were associated with each magnitude of each reinforcer. Cocaine choice was directly related to cocaine dose, and a cocaine choice dose-effect curve could be determined in a single experimental session. The choice between cocaine and food was influenced by the schedules of cocaine and food reinforcement, the magnitude of the food reinforcer, and the amount of noncontingent food provided outside the experimental session. These results confirm and extend previous findings with other choice procedures and validate the sensitivity of the present procedure to environmental manipulations. The choice between cocaine and food could also be influenced by treatment with candidate pharmacotherapies for cocaine abuse and dependence. The "agonist" medication d-amphetamine produced rightward shifts in the cocaine choice dose-effect curve and decreased cocaine choice, whereas the "antagonist" medication flupenthixol had little effect on cocaine choice. Overall, these results suggest that this choice procedure may be useful for the evaluation of both environmental determinants of cocaine use and candidate pharmacotherapies for the treatment of cocaine abuse.