To determine the immediate effects of intravenous "recreational" doses of cocaine on myocardial ventricular relaxation and contraction and on coronary blood flow. To determine the cardiac effects of cocaine after the administration of propranolol, as propranolol has been used to limit the cardiovascular effects of cocaine. Prospective study. Twenty mongrel dogs. We continuously recorded central aortic pressure, left atrial and ventricular pressures, coronary artery blood flow, and electrocardiograms in each dog. We determined from the left ventricular pressure waveforms the maximum rate of pressure increase [(dP/dt)max] and the time constant of isovolumic ventricular relaxation as our indices of ventricular contraction and relaxation. In our initial series of experiments, we obtained pressure, coronary artery blood flow, and electrocardiographic recordings in ten anesthetized dogs before and for 40 mins after the intravenous administration of cocaine, in doses of 2.5 and then 5 mg/kg. In our second series of experiments in ten additional dogs, we injected 0.5 mg/kg of propranolol intravenously 30 mins before the injection of cocaine (2.5 mg/kg), and obtained hemodynamic and electrocardiographic recordings before and for 40 mins after the injection of propranolol and cocaine. Cocaine, 2.5 mg/kg, abruptly increased the time constant of isovolumic ventricular relaxation from 22.9 +/- 1.2 to 29 +/- 2.2 msecs at 1 min (p < .05) and to 35.3 +/- 2 msec at 40 mins (p < .01) but did not significantly change the mean arterial pressure, left atrial pressure, heart rate, coronary blood flow, or the maximum rate of left ventricular pressure increase [(dP/dt)max]. Cocaine also progressively displaced the electrocardiographic ST segments by 3.2 +/- 0.6 mm (p < .01) over 40 mins. Cocaine, 5 mg/kg, rapidly increased the time constant of isovolumic ventricular relaxation from 28.5 +/- 2.5 to 41 +/- 3 msecs in 1 min (p < .05) and to 48.7 +/- 4 msecs at 40 mins (p < .01) and reduced (dP/dt)max from 2905 +/- 370 to 1422 +/- 121 mm Hg/sec at 1 min (p < .01); (dP/dt)max returned to 2351 +/- 415 mm Hg/sec during the next 39 mins. Cocaine did not significantly change either the mean arterial or left atrial pressures. However, this dose of cocaine did decrease, over 40 mins, the heart rate from 184 +/- 11 to 139 +/- 11 beats/min (p < .01) and reduced coronary blood flow by 20% (p < .01). Cocaine also displaced the electrocardiographic ST segments by 3.3 mm over 40 mins (p < .05). Cocaine and propranolol abruptly increased the time constant of isovolumic ventricular relaxation from 26.4 +/- 1.3 to 43.2 +/- 2.1 msecs (p < .01) at 1 min and to 46.8 +/- 1.5 msecs at 3 mins (p < .01). The time constant of isovolumic ventricular relaxation remained abnormally increased at 43.0 +/- 1.4 msecs at 40 mins. Cocaine and propranolol reduced (dP/dt)max from 2760 +/- 458 mm Hg/sec to a minimum value of 1400 +/- 119 mm Hg/sec at 2 mins (p < .01). However, (dP/dt)max then returned to 2201 +/- 359 mm Hg/sec during the next 38 mins. Cocaine and propranolol did not significantly change the mean arterial and left atrial pressures, or heart rate, but did reduce coronary blood flow, over 40 mins, by 25% (p < .001). Cocaine also maximally displaced the electrocardiographic ST segments by 1 +/- 0.2 mm (p < .01). Cocaine substantially impairs myocardial ventricular relaxation for periods of at least 40 mins. Propranolol significantly intensifies cocaine's depressant effect on ventricular relaxation.