The present study examined the effects of two novel dopamine D3 receptor compounds, NGB 2904 [N-(4-(4-(2,3-dichlorophenyl)piperazin-1-yl)butyl)-9H-fluorene-2-carboxamide], an antagonist, and CJB 090 [N-(4-(4-(2,3-dichlorophenyl)piperazin-1-yl)butyl)-4-(pyridin-2-yl)benzamide], a partial agonist, in two models of cocaine abuse in rhesus monkeys. To establish a dose range and time course of effects, both compounds were shown to block quinpirole-induced yawning when administered i.m. 15, 30, or 120 min before quinpirole. Next, rhesus monkeys were trained to discriminate i.m. injections of saline (0.5 ml) and cocaine (0.3 mg/kg). Neither D3 compound (0.03-3.0 mg/kg; n=3) substituted for cocaine in any monkey. When given in combination with cocaine, CJB 090 but not NGB 2904 attenuated the discriminative stimulus effects of cocaine, shifting the cocaine dose-response curve to the right. In a separate group of monkeys, responding was maintained under a second-order schedule of either food (1.0-g pellets; n=3) or cocaine (0.1 mg/kg/injection; n=4) presentation. When responding was stable, a dose of NGB 2904 (1.0-5.6 mg/kg i.v.) or CJB 090 (0.3-3.0 mg/kg i.v.) was administered for 5 consecutive days, immediately before the session. CJB 090, but not NGB 2904, decreased cocaine- and food-maintained responding. These data indicate that compounds with relatively high affinity and selectivity for the D3 receptor can attenuate the discriminative and reinforcing stimulus effects of cocaine while not producing cocaine-like effects. The present findings support the continued examination of D3 compounds as pharmacological tools for better understanding the role of this receptor subtype in cocaine addiction and as potential lead compounds for novel therapeutic agents.