It has been suggested that toxicity to cocaine is related to the relative rate of cocaine metabolism by cholinesterases and to activation of cholinergic receptors either directly or by reflex mechanisms. We examined these possibilities by altering cholinesterase activity and blocking cholinergic receptors in rats prone or resistant to cocaine-induced cardiovascular toxicity. Rats were instrumented with a pulsed Doppler flow probe on the ascending aorta for measurement of cardiac output and cannulated for arterial pressure and heart rate determination. In conscious rats, cocaine (5 mg/kg iv) elicited pressor responses and a delayed bradycardia but cardiac output and systemic vascular resistance responses varied greatly between rats. Pretreatment with the nonspecific cholinesterase inhibitors physostigmine (0.1-0.2 mg/kg) or neostigmine (0.1 mg/kg) reduced the pressor response by diminishing the increase in systemic vascular resistance. In contrast, inhibition of cocaine metabolism with the selective plasma cholinesterase inhibitor tetraisopropyl pyrophosphoramide (0.5 mg/kg) or increasing cholinesterase activity with human butyryl cholinesterase (9.9 mg/kg iv) did not alter hemodynamic responses to cocaine. Administration of atropine methyl bromide (0.5-1 mg/kg iv) alone or with physostigmine to prevent the cholinomimetic effects of physostigmine reduced the cocaine-induced decrease in cardiac output noted in some animals. These data suggest that the cocaine-induced decrease in cardiac output observed in some rats is, at least in part, dependent on activation of muscarinic receptors. In addition, the rate of cocaine metabolism is not critical for the initial hemodynamic responses to cocaine in conscious rats.