Cocaine is a potent sympathomimetic drug that can provoke lethal cardiac events. Cocaine-induced alterations in autonomic balance, particularly during myocardial ischemia, could contribute significantly to these adverse reactions. To test this hypothesis, we produced a 2-min left circumflex coronary artery (LCX) occlusion in unanesthetized mongrel dogs (n = 7) instrumented to measure left ventricular pressure (LVP), ventricular electrogram, and coronary blood flow (CBF) with and without various doses of cocaine (0.0, 0.5, 1, 2, and 4 mg/kg). At least 24 h elapsed between cocaine doses, which were given in random order. Time series analysis of heart rate (HR) variability was used as an index of cardiac vagal tone (0.24-1.04 Hz). Cocaine elicited dose-dependent increases in HR that were accompanied by corresponding decreases in cardiac vagal tone. The peak response was achieved approximately 1 min after cocaine was given and returned to precocaine values 15 (0.5 and 1 mg/kg), 30 (2 mg/kg), or 60 (4 mg/kg) min later. Myocardial ischemia elicited significant increases in HR and reductions in cardiac vagal tone that were accentuated by cocaine (1, 2, and 4 mg/kg); e.g., cocaine (2 mg/kg) elicited a greater HR (control 119.3 +/- 5.9, occlusion 149.7 +/- 9.6; cocaine 144 +/- 11.9, occlusion 178.3 +/- 10.4 beats/min) and vagal tone (control 5.6 +/- 0.7, occlusion 2.6 +/- 0.3; cocaine 5.2 +/- 0.7, occlusion 1.3 +/- 0.5 ln s2) response to 2-min coronary occlusion. beta-Adrenoceptor blockade (propranolol HCl 1 mg/kg) attenuated the HR response but elicited greater reduction (lower values were achieved) in vagal tone during coronary artery occlusion.(ABSTRACT TRUNCATED AT 250 WORDS)