Cocaine administration elicits a pressor response reportedly dependent upon central and peripheral actions. The present study was designed specifically to determine whether cocaine affects baroreflex sensitivity and whether this is mediated by an action in the CNS. Baroreflex control of heart rate was assessed by noting the change in the duration of the arterial pulse pressure or in the heart rate elicited by administration of pressor and depressor agents. The effects of intracerebroventricular and intravenous cocaine administration were compared. Cocaine (0.1-0.15 mg/kg/min, i.v.) produced a decrease in baroreflex sensitivity that was greater than that produced by an approximately equipressor infusion of phenylephrine (0.5-1.3 micrograms/kg/min, i.v.). Intracerebroventricular (i.c.v.) administration of cocaine (1.5-150 micrograms) or procaine (100 micrograms) had no effect on arterial pressure, heart rate or baroreflex control of heart rate. Furthermore, central administration of yohimbine (3 and 10 micrograms, i.c.v.) was capable of preventing the suppression of the heart rate responses induced by cocaine administration. Propranolol (15 micrograms, i.c.v.) was not able to attenuate the suppression of baroreflex sensitivity elicited by intravenous cocaine administration but higher doses (50-150 micrograms, i.c.v.) could mimic this effect. These data suggest that cocaine suppresses baroreflex control of heart rate by a central alpha 2-adrenergic mechanism. In contrast, the pressor response to intravenous cocaine is not likely to be dependent upon a forebrain periventricular site.