Many of the behavioral effects of cocaine are thought to be mediated via dopamine (DA) reuptake inhibition and a subsequent increase in DA activity. As a DA agonist with selectivity for autoreceptors, PD 128483 has behavioral effects consistent with decreased DA neurotransmission and might be expected to attenuate the behavioral effects of cocaine. The present study was designed to evaluate PD 128483 {(+)-4,5,5a,6,7,8-hexahydro-6-methyl-thiazolo[4,5-f]quinoline-2-amine} alone and in combination with cocaine in three behavioral paradigms. One group of rhesus monkeys (n = 7) was allowed to self-administer cocaine (0.03 or 0.1 mg/kg/injection i.v., fixed-ratio 10, 2 hr/day) in baseline sessions. When behavior was stable, PD 128483 (0.01-0.3 mg/kg/injection i.v.) was made available for self-administration, and failed to maintain responding in three of four monkeys tested. Po 128483 (1.0-10.0 mg/kg i.m.) decreased cocaine self-administration in all four monkeys tested when given before cocaine availability. Another group of monkeys (n = 4) was trained to discriminate cocaine (0.2 or 0.4 mg/kg i.m.) from saline in a two-lever, food-reinforced drug discrimination paradigm. High doses of PD 128483 (3.2-12.8 mg/kg i.m.) substituted for cocaine as a discriminative stimulus in three of four monkeys. Administered in combination with cocaine, PD 128483 had little effect on the cocaine discriminative stimulus up to doses which decreased response rate. PD 128483 (1.0-2.0 mg/kg i.p.) decreased milk intake in rats (n = 9) and, when given in combination with cocaine, shifted the cocaine dose-response function (4.0-16.0 mg/kg i.p.) to the left. Thus, the behavioral effects of PD 128483 in combination with cocaine were not consistent with selective autoreceptor actions but may be the result of partial agonist actions at postsynaptic DA D2 receptors. Drugs like Po 128483 that may have some cocaine-like subjective effects but relatively low liability for abuse may prove useful in the treatment of cocaine abuse.