This study tests the hypothesis that more muscle salvage after acute ischemia is possible by "gentle," temporary reperfusion than with sudden, complete revascularization. Ten dogs underwent 4 hours of left anterior descending coronary artery ligation with reperfusion during total vented bypass for 1 hour of the 2-hour reperfusion period. In five dogs, reperfusion was accomplished by release of the occlusion suddenly and completely. The five others received selective low-pressure (40 to 50 mm Hg) coronary reperfusion with normal blood for 20 minutes at 30 ml/min before the occlusion was relieved completely. Systolic shortening with ultrasonic crystals, triphenyltetrazolium chloride staining, and myocardial wet and dry weights were measured. Sudden relief of occlusion failed to restore contractility spontaneously (-7 +/- 1% systolic shortening, p less than 0.05) or with inotropic infusion (-2 +/- 4% systolic shortening, p less than 0.05) and caused the greatest amount of edema (82.2%, systolic shortening, p less than 0.05) and triphenyltetrazolium chloride nonstaining (76% area at risk, p less than 0.05). In contrast, temporary, gentle reperfusion allowed slight spontaneous recovery in four of five hearts (4 +/- 2% systolic shortening), increasing to 26 +/- 12% systolic shortening (p less than 0.05) with inotropic stimulation, limited edema (80.7%, p less than 0.05), and reduced triphenyltetrazolium chloride nonstaining to 55% (p less than 0.05). Early temporary, gentle reperfusion limits the postischemic damage that occurs with sudden, complete revascularization (aortic unclamping without control of reperfusion pressure or flow). These findings may have implications during revascularization for acute myocardial infarction when perfusion pressure and flow can be controlled.