Superoxide dismutase (SOD) has been documented to limit myocardial infarct size in the richly collateralized dog heart. This study was designed to explore this concept in a low-collateralized animal model. A blind, randomized, placebo-controlled protocol was used in 65 pentobarbital-anesthetized pigs subjected to closed-chest left anterior descending coronary artery occlusion for 30 (n = 22), 60 (n = 22), and 90 (n = 14) minutes followed by reperfusion up to 24 hours from the start of occlusion. Another seven control pigs were subjected to 24 hours of permanent occlusion. A total dose of 9 mg/kg bovine CuZn SOD was administered as a bolus injection immediately before reperfusion followed by a 1-hour infusion. Infarct size was assessed by tetrazolium staining. Myocardium at risk and collateral flow were determined by using cerium-141-labeled microspheres (15 microns) during the occlusion. After 30 minutes of occlusion, infarct sizes in placebo versus SOD-treated animals were 45.5 +/- 15.7% vs. 23.8 +/- 15.6% of myocardium at risk (p = 0.007). The corresponding values after 60 minutes of occlusion were 78.6 +/- 9.3% vs. 66.9 +/- 14.6% (p = 0.035). SOD administered after 90 minutes of occlusion did not limit infarct size (88.5 +/- 4.8% vs. 92.3 +/- 5.2%). Twenty-four hours of coronary occlusion resulted in infarction of 92.4 +/- 4.2% of myocardium at risk. (All values are mean +/- SD.) Ventricular fibrillation occurred in only nine pigs distributed equally between SOD and placebo. The results indicate that CuZn SOD has the potential to further improve the myocardial salvage established by reperfusion of an ischemic pig heart territory. However, the narrow time window for limiting infarct size in the pig by reperfusion is not much extended by SOD.