Recent studies suggest that oxygen-derived free radicals contribute to the pathogenesis of postischemic myocardial dysfunction (myocardial "stunning"). This concept, however, is predicated exclusively on results obtained in open-chest preparations, which are subject to the confounding influence of many unphysiological conditions. The lack of supporting evidence in more physiological animal models represents a major persisting limitation of the oxy-radical hypothesis of myocardial stunning. The goal of this study was to address two fundamental (and related) questions: 1) Does the open-chest animal model alter the phenomenon of myocardial stunning? 2) If so, how valid are the concepts, derived from such a model, regarding the pathogenetic role of oxy-radicals? In part 1 of the study, myocardial stunning after a 15-minute coronary occlusion was compared in 30 pentobarbital-anesthetized open-chest dogs and in 19 conscious dogs. For any given level of collateral flow during occlusion, the recovery of systolic wall thickening after reperfusion was markedly less in open-chest animals. In an additional group of five open-chest dogs, a close inverse relation was noted between body temperature and postischemic wall thickening, indicating that the recovery of the stunned myocardium in acute experiments may vary markedly depending on how temperature is controlled. Because of these major differences between open-chest and conscious dogs, the oxy-radical hypothesis needs to be tested in the latter model. Thus, in part 2 of the study, conscious unsedated dogs undergoing a 15-minute coronary occlusion were randomized to an intravenous infusion of either saline (19 coronary occlusions) or superoxide dismutase (SOD) plus catalase (CAT) (21 coronary occlusions). Despite the fact that the plasma levels of SOD and CAT declined rapidly after reperfusion, postischemic wall thickening was significantly greater in treated compared with control dogs throughout the first 6 hours of reflow. Thus, a brief (60-minute) infusion of SOD and CAT produced a sustained improvement of recovery of contractility. The magnitude of this beneficial effect was a function of the severity of ischemia: the lower the collateral perfusion, the greater the improvement effected by the enzymes. The accelerated recovery produced by SOD and CAT was not followed by any deterioration of contractility, suggesting that postischemic dysfunction is not a teleologically "protective" phenomenon. In conclusion, the severity of myocardial stunning is greatly exaggerated by the unphysiological conditions present in the barbiturate-anesthetized open-chest dog.(ABSTRACT TRUNCATED AT 400 WORDS)