Recent evidence suggests that postischemic myocardial dysfunction (or myocardial "stunning") may be mediated by oxygen free radicals, but the mechanism for their production remains unknown. To explore the role of xanthine oxidase as a potential source of free radicals, open-chest dogs undergoing a 15-min occlusion of the left anterior descending coronary artery (LAD) followed by 4 h of reperfusion (REP) received intravenously either allopurinol (50 mg/kg 48 h, 20 h, and 30 min before occlusion, 10 mg/kg 1 min before REP, and 6.25 mg X kg-1 X h-1 throughout REP, n = 13) or saline (n = 14). The two groups were similar with respect to occluded bed size (postmortem perfusion) and collateral flow (radioactive microspheres). In controls, the transcardiac difference in plasma uric acid (great cardiac vein - arterial concentration) increased 199 +/- 70% (means +/- SE) during ischemia (P less than 0.02) and remained elevated for 5 min after REP; no increase was observed in treated dogs. Regional myocardial function was assessed by measuring systolic wall thickening with an epicardial Doppler probe. The two groups exhibited comparable systolic thickening under base-line conditions and similar degrees of dyskinesis during ischemia. Following REP, however, recovery of contractile function (expressed as percent of preocclusion values) was considerably greater in allopurinol-treated as compared with control dogs: 57 +/- 14 vs. -22 +/- 16 (P less than 0.01) at 1 h, 70 +/- 13 vs. -15 +/- 15 (P less than 0.001) at 2 h, 65 +/- 14 vs. -28 +/- 13 (P less than 0.001) at 3 h, and 68 +/- 13 vs. -17 +/- 14 (P less than 0.001) at 4 h. These differences could not be ascribed to hemodynamic factors. The results suggest that xanthine oxidase is a source of the oxygen free radicals responsible for myocardial stunning following a brief episode of reversible regional ischemia.