Hydroxyl radical formation, secondary to superoxide radical generation, has been advocated as the actual mechanism of oxygen radical-mediated damage in biological systems. The present study was designed to compare the efficacy of administration of the hydroxyl radical scavenger mannitol vs. that of the superoxide radical scavenger superoxide dismutase (SOD) in reducing myocardial reperfusion injury, and to test whether combined treatment with both agents would confer better tissue protection compared with either intervention alone. Rabbit hearts perfused within a ³¹P nuclear magnetic resonance (³¹P-NMR) spectrometer were subjected to 30 minutes of total global ischemia at 37°C. At reflow, 12 hearts in each group received either (a) a bolus of standard perfusion buffer, followed by 45 minutes of reperfusion (controls); (b) the superoxide radical scavenger recombinant human SOD (h-SOD, as a 60,000 U bolus followed by a 100 U/ml infusion for 15 minutes); (c) the hydroxyl radical scavenger mannitol (50 mM bolus followed by 15 minutes of 50 mM infusion); or (d) a combination of both agents. All treated hearts were switched to standard buffer for the remaining 30 minutes of reperfusion. Treatment with h-SOD alone was associated with a significant improvement in the recovery of cardiac contractility and coronary flow, as well as of ATP content, compared to control hearts. In contrast, mannitol treatment resulted in a small, nonsignificant improvement in these parameters. The addition of mannitol to h-SOD did not result in further significant improvement of contractility and ATP recovery compared to h-SOD alone. These data demonstrate that under our experimental conditions significant protection against reperfusion injury can be achieved by the administration of h-SOD alone, without the need for additional hydroxyl radical scavenger therapy with mannitol. These results do not exclude that significant tissue protection may be achieved by different doses of mannitol or by other agents. However, they suggest that under definite experimental conditions prevention of hydroxyl radical formation, rather than attempts to minimize hydroxyl radical toxicity, might be a more efficient method to prevent oxygen radical-mediated reperfusion injury in isolated hearts.