We examined 20 patients undergoing coronary bypass grafting for coronary artery disease with NYHA classifications of II and III who had been treated with beta-blocking agents. Patients were randomised for administration of either adrenaline (0.1 microgram/kg/min) or amrinone (bolus 1 mg/kg, continuous infusion of 5-10 micrograms/kg/min), if following cardiopulmonary bypass their cardiac index was < 2.4 L/min/m2 with normal peripheral resistance and normal or increased right- or left-ventricular filling pressures. Over a period of 1 hour, the hemodynamic parameters mean arterial pressure (MAP), cardiac index (CI), heart rate (HR), coronary perfusion pressure (CPP), total peripheral resistance (TPR), as well as the pressure-work index (PWI) were registered or calculated. By means of a coronary sinus catheter myocardial arterio-venous oxygen content difference (AVDO2cor), myocardial blood flow (MBF), using the thermodilution method, and myocardial oxygen consumption (MVO2) could be measured or calculated. Simultaneously, arterial and myocardial lactate concentrations and, using the arterio-venous lactate ratio, myocardial lactate extraction or production were quantified. Using a transseptal approach, the left-ventricular pressure curve was measured and used to differentiate for myocardial contractility (dp/dtmax). Following induction of anesthesia and after cardiopulmonary bypass, plasma levels of the used beta-blocking agent were determined. Both substances caused a significant increase in myocardial contractility, with adrenaline showing a more potent effect than amrinone. Both substances caused a significant increase in CI with a mild increase in HR. Amrinone caused a significant drop in TPR, while MAP remained practically constant.(ABSTRACT TRUNCATED AT 250 WORDS)