Hibernating myocardium has been defined as a persistent impairment of contractile function resulting from reduced coronary blood flow that can be partially or completely resolved once coronary perfusion is restored. In fact, recent clinical reports have documented a dramatic improvement in contractile function after relief of chronic sublethal ischemia. To investigate the phenomenon of sublethal ischemia followed by reperfusion, we assessed myocyte morphology, high-energy phosphate content, and regional contractile function in dogs undergoing (1) 3 hours of subtotal coronary artery occlusion (CO) and 3 hours of reflow or (2) 3 hours of total CO followed by reflow, in which myocyte viability was maintained by extensive collateral perfusion during ischemia (total CO/negligible necrosis). Data were compared with findings in a third group of dogs with total CO and low collateral blood flow during ischemia, in which large confluent infarcts developed. Endocardial blood flow averaged 30 ± 6% (p < 0.01) and 27 ± 4% (p < 0.01) of baseline preocclusion values during ischemia in the groups with subtotal CO and total CO/negligible necrosis, versus 3 ± 1% of baseline values in dogs with total CO/confluent necrosis. Both the subtotal CO and total CO/negligible necrosis groups exhibited only mild-to-moderate reversible myocyte injury (assessed by electron microscopy) and had essentially no necrosis: infarct size was 1 ± 1% (p < 0.01) and 4 ± 2% (p < 0.01) of the risk region in the subtotal CO and total CO/negligible necrosis groups, versus 55 ± 9% of the risk region in the total CO/confluent necrosis group. Furthermore, myocardial high-energy phosphate stores were in part preserved in all dogs that underwent sublethal ischemia: endocardial adenosine triphosphate (ATP) content was 55 ± 11% (p < 0.01) and 56 ± 8% (p < 0.01) versus 11 ± 2% of baseline values in the subtotal CO, total CO/negligible necrosis, and total CO/confluent necrosis groups, respectively. At 3 hours post occlusion, segment shortening averaged +21 ± 10% of baseline values in dogs with subtotal CO, (p <0.01 versus both total CO groups), -29 ± 9% in dogs with total CO/negligible necrosis, and -36 ± 13% in dogs with total CO/confluent necrosis. Reperfusion after sublethal ischemia produced an acute improvement in contractile function in both the subtotal CO and total CO/negligible necrosis groups. This functional rebound was short lived, however; by 3 hours after reflow, segment shortening had deteriorated to values that did not differ from those measured during ischemia (+20 ± 15% and -24 ± 5% of baseline values in dogs with subtotal CO and total CO/negligible necrosis, respectively). Thus a 3-hour episode of sublethal ischemia in this canine model results in a reduction in contractile function during hypoperfusion associated with preserved myocardial high-energy phosphate stores and prevention of myocyte necrosis. However, in contrast to some clinical reports, reperfusion after sublethal ischemia was not associated with a dramatic improvement in contractile function. Rather the postischemic dysfunction, partial preservation of ATP stores, and creatine phosphate overshoot observed after relief of sublethal ischemia suggest that this previously hypoperfused but viable tissue is stunned during the initial hours after reflow.