Buspirone and gepirone were evaluated as potential pharmacotherapies for cocaine abuse by studying the effects of acute and repeated treatment on IV cocaine self-administration in rhesus monkeys. Chlorpromazine was also evaluated as a positive control. Effects of IV drug pretreatments were tested during daily 60-min sessions with lever-pressing reinforced under a fixed-ratio 10 schedule of 0.02 or 0.05 mg/kg cocaine infusions. Acute pretreatment with buspirone (0.1 and 0.3 mg/kg, IV) increased rates of cocaine self-administration without disrupting food pellet consumption. Some doses of buspirone also produced changes in rates of cocaine self-administration without altering the within-session pattern of responding. In contrast, acute doses of gepirone had little effect on rates of cocaine self-administration, while disruptions in food consumption and changes in the within-session pattern of cocaine self-administration were obtained at the highest dose of gepirone tested (1.0 mg/kg). When either buspirone (0.1 and 0.3 mg/kg, IV) or gepirone (0.1 mg/kg, IV) were administered daily for 10 days, consistent effects on cocaine self-administration were not observed. Thus, the effects of acute buspirone administration on cocaine-maintained behavior were similar to the effects produced by chlorpromazine and other dopaminergic antagonists, whereas, gepirone was ineffective. These results provide some support for further evaluation of buspirone as a potential pharmacotherapy for cocaine abuse, although its lack of efficacy with repeated treatment is not encouraging. The negative results with gepirone provide less rationale for continued investigations with this drug, possibly because of its lesser involvement than buspirone with dopaminergic neurotransmission.