To characterize the functional and biochemical recovery of myocardium salvaged by reperfusion, 19 anesthetized mongrel dogs underwent 2 hour occlusion of the proximal left anterior descending coronary artery, followed by reperfusion for up to 2 weeks. Thirteen dogs had a permanent occlusion and served as the control group. All dogs had serial two-dimensional echocardiograms and in vivo biopsies for adenosine triphosophate (ATP) and creatine phosphate after occlusion and at the time of sacrifice (6 hours, n = 15; 7 days, n = 6; or 14 days after occlusion, n = 11). The area of necrosis and area at risk of necrosis were identified in dogs sacrificed at 6 hours. After 4 hours of reperfusion, the area of necrosis determined by the triphenyltetrazolium chloride technique, expressed as a function of the area at risk by in vivo monastral blue injection, was 23.9 ± 5.8% (mean ± standard error of the mean) and largely subendocardial, compared with 89.1 ± 5.3% in dogs with permanent (6 hour) occlusion (probability [p] < 0.001), in which it was transmural. Myocardial contractility was assessed by measuring systolic wall thickening by computer-assisted two-dimensional echocardiography. Before occlusion, this variable averaged 47.9 ± 6.3% in the region to become ischemic. An early salutary effect of reperfusion in the central ischemic zone was seen as a decrease in wall thinning from -11.9 ± 2.9%, 90 minutes after occlusion, to -7.4 ± 2.0%, 4 hours after reperfusion, contrasted with a change from -6.2 ± 3.3%, 90 minutes after occlusion, to -11.7 ± 3.3%, 6 hours after occlusion in the permanently occluded group (p < 0.05). Active thickening was not observed until 72 hours of reperfusion (13.1 ± 4.6%); this increased to 19.5 ± 2.8% at 14 days, compared with 1.5 ± 5.9% in the permanently occluded group at this time (p < 0.05). However, preocclusion values were not achieved in the reperfused group. Some recovery of ATP in the salvaged subepicardium was also noted, with return of levels from 8.0 ±1.9 nmol/mg protein, 90 minutes after occlusion, to 13.2 ± 1.8 nmol/mg protein, after 4 hours of reperfusion (p < 0.05); however, 7 days were required for recovery approaching preocclusion values of 34.9 ± 1.9 nmol/mg protein. Thus, reperfusion after 2 hours of coronary occlusion resulted in recovery of jeopardized myocardium, in which the deterioration of function observed in nonreperfused tissue was reversed and biochemical improvement occurred over the course of 2 weeks. Computer-assisted two-dimensional echocardiography, a noninvasive but quantitative technique, appears well suited to monitor the rate of functional recovery of myocardium salvaged by reperfusion.