Cocaine (5-40 mg/kg, intraperitoneally) enhanced locomotion and rearing accompanied with head circling and body shaking. Although at 40 mg/kg typical stereotypy licking occasionally appeared, 40% of the rats died. At doses that did not affect physiologic locomotion and rearing, the D-1-receptor antagonist SCH23390 but not D-2 antagonist raclopride inhibited locomotion and rearing stimulated by cocaine (20 mg/kg). All behavioral responses of cocaine were abolished with increasing doses of raclopride and SCH23390. Sulpiride, a D-2 antagonist, exerted a biphasic effect on locomotor activity (i.e., a low dose of sulpiride increased and a high dose decreased cocaine-induced locomotor activity). Sulpiride enhanced head circling, body shaking, and increases of rearing induced by cocaine. D-2-receptor agonists quinpirole and bromocriptine inhibited these responses, presumably by activating the typical stereotyped behaviors such as sniffing at low doses, and licking and gnawing at high doses. The lowest dose of bromocriptine inhibited all behaviors induced by cocaine without producing typical stereotyped behaviors in itself. SK&F38393, a D-1-receptor agonist, in combination with cocaine did not induce typical stereotypy, which results in a synergistic effect of D-1 and D-2-receptor activities. The increases of locomotion and rearing, head circling, and body shaking induced by cocaine may involve the indirect activation of postsynaptic D-1 and D-2 receptors, presumably via dopamine release, resulting from inhibition of the presynaptic D-2 receptors. These results also provide evidence that the indirect stimulation of postsynaptic D-2 receptors by cocaine (20 mg/kg) is insufficient to induce stereotyped behaviors, and that the role of dopamine D-1 receptors in mediating the behavioral actions of acute cocaine appears to be more important than that of D-2 receptors. Our results also suggest that bromocriptine may be useful for the treatment of acute cocaine poisoning.