Recent evidence suggests that cocaine decreases rather than increases sympathetic nerve discharge (SND). The purpose of the present study was to provide the first complete characterization of the dose-response relationships of cocaine (0.005-3 mg/kg, IV) for arterial pressure, heart rate, and lumbar, splanchnic, or renal SND in pentobarbital-anesthetized rats. Cocaine was also tested in conscious rats. In pentobarbital-anesthetized rats cocaine elicited prolonged (lasting up to 56 min), dose-dependent decreases in SND on all three nerves. The splanchnic nerve was significantly more sensitive to the inhibitory actions of cocaine than was the lumbar nerve. Cocaine increased arterial pressure and elicited bradycardia at doses above 0.5 mg/kg. Comparison of the dose-response curves of cocaine for splanchnic SND in sham-operated and sinoaortically deafferentated (SAD) rats showed that the baroreceptor reflex made only a minor contribution to the magnitude of sympathoinhibitory response. However, the duration of the sympathoinhibitory response was significantly shorter in SAD than in sham animals. In conscious rats, cocaine (0.1 and 1.0 mg/kg) elicited a pattern of neural and cardiovascular responses similar to that seen in anesthetized rats, except that the prolonged sympathoinhibitory responses were preceded by a brief (lasting < 10 s) increase in SND. From these data we conclude that cocaine produces prolonged decreases in SND in conscious and anesthetized rats. These sympathoinhibitory responses do not appear to result from baroreceptor reflex activation and may involve a central mechanism of action.