Ten dogs underwent 45 minutes of normothermic ischemic arrest. After 15 minutes of reoxygenation, none could support the systemic circulation independently. In five dogs, we could discontinue bypass (cardic output 70 to 100 cc/kg/min) by giving dopamine (10 to 30 mcg/kg/min). In five other dogs, total cardiopulmonary bypass was prolonged for an additional 30 minutes and no dopamine was given. During control and at 15 and 45 minutes after aortic unclamping, we measured myocardial blood flow (microspheres), metabolism (oxygen uptake and lactate), water content (wet/dry weight), left ventricular compliance (intraventricular balloon), and performance (isovolumetric and Starling function curves). Dogs treated with prolonged bypass showed progressive improvement in ventricular compliance, function, and water content and did not require inotropic drugs when bypass was discontinued 45 minutes after ischemia. In contrast, dogs receiving dopamine exhibited more myocardial edema (3.3% versus 1.7% water gain), worse ventricular compliance (18% versus 55% recovered at 25 ml left ventricular volume), poorer contractility (58% versus 70% recovery of +dP/dt), generated 50% less stroke work at a left atrial pressure of 25 mm Hg (0.25 versus 0.52 gm/kg), failed to augment oxygen uptake to meet the metabolic demands of the working heart (11% versus 45% increase in oxygen uptake), and required continued inotropic support to discontinue extracorporeal circulation. We conclude that (1) limited prolongation of total bypass enhances recovery from ischemic damage and (2) use of inotropic drugs to prematurely discontinue extracorporeal circulation will impede recovery by accentuating myocardial edema and further decreasing ventricular compliance, performance, and oxygen utilization.