The mechanism for the prolonged contractile dysfunction observed in myocardium reperfused after reversible regional ischemia ("stunned" myocardium) is unclear. Recent studies suggest that myocardial stunning may be mediated by oxygen-derived free radicals, but the precise molecular species involved remain unknown. Thus we explored the role of the highly cytotoxic hydroxyl radical in regional postischemic dysfunction by using dimethylthiourea (DMTU), an effective and highly permeable hydroxyl radical scavenger. Open-chest dogs undergoing a 15 min occlusion of the left anterior descending coronary artery followed by 4 hr of reperfusion received either DMTU (0. 5 g/kg iv over 45 min starting 30 min before occlusion, n = 14) or saline (n = 15). Control and treated dogs were comparable with respect to variables that may affect postischemic dysfunction, including heart rate, aortic pressure, left atrial pressure, arterial blood gases and hemoglobin concentration, size of the occluded bed (determined by postmortem perfusion), and collateral blood flow (determined by radioactive microspheres). Regional myocardial function was assessed by measuring wall thickening with an epicardial Doppler probe. The two groups exhibited comparable systolic thickening under baseline conditions and similar degrees of dyskinesis during ischemia. After reperfusion, however, wall thickening (expressed as percent of baseline) was considerably greater in treated as compared with control dogs: 53 + 9% (mean -+- SEM) vs 9 ± 14% (p < .03) at 1 hr, 55 ± 9% vs 23 ± 13% (p < .05) at 2hr, 60 + 9%vs28 ± 14% (p < .05) at 3hr, and67 ± 5%vs36 + 13% (p < .05) at 4hr. Thus DMTU produced a significant and sustained improvement in recovery of contractile function. In concentrations greater than the plasma levels attained in vivo, DMTU did not scavenge either hydrogen peroxide or superoxide anion in vitro. These results suggest that the myocardial dysfunction occurring after a brief episode of regional ischemia is mediated in part by the hydroxyl radical.