Controversy persists regarding which oxygen metabolites are cytotoxic. Although the combination of superoxide dismutase (SOD) and catalase has been shown to attenuate postischemic myocardial dysfunction ("stunning"), it is unknown whether this beneficial effect is due to scavenging of O2-., H2O2, or both. Accordingly, 85 open-chest dogs underwent a 15-min occlusion of the left anterior descending coronary artery followed by 4 h of reperfusion. In phase A, dogs received an intravenous infusion of saline (group I), SOD (5 mg/kg, group II), catalase (12,000 U/kg, group III), or the combination of SOD and catalase (same doses, group IV). Recovery of regional myocardial function (assessed as systolic wall thickening) after reperfusion was significantly improved by the combination of SOD and catalase but not by SOD alone or catalase alone. To determine whether higher doses of enzymes are more effective, in phase B dogs received an intracoronary infusion of normal saline (group V), SOD in low dose (1.5 mg/kg, group VI), SOD in high dose (6.3 mg/kg plus 1.5 mg/kg iv, group VII), catalase in low dose (18,000 U/kg, group VIII), or catalase in high dose (240,000 U/kg plus 40,000 U/kg iv, group IX). Despite the fact that the local plasma levels of enzymes were considerably higher than those achieved in phase A, none of the treatments in phase B significantly enhanced recovery of contractile function. This study demonstrates that the combination of SOD and catalase is more effective than either enzyme alone in attenuating postischemic myocardial dysfunction and that increasing the doses of SOD or catalase does not provide additional protection. The results suggest that both O2-. and H2O2 contribute significantly to the pathogenesis of myocardial stunning after regional ischemia in the intact animal. Furthermore, the data imply that if SOD and catalase are to be used clinically to prevent postischemic dysfunction, protection may be achieved most effectively by combining the two enzymes.