Myocardial reperfusion injury may be due to biophysical changes (e.g., endothelial cell junctional separations), as well as biochemical mechanisms (e.g., oxygen free radical activity). Superoxide dismutase (SOD), a free radical scavanger, may be effective in reducing chemical injury. Fractions of hydroxyethyl starch (HES-Pz), a large macromolecule, have been shown to decrease microvascular permeability associated with reperfusion-induced biophysical alterations. A comparison of SOD to HES-Pz was performed using a canine model of 1-hour left anterior descending coronary artery (LAD) clamping followed by 24 hours of reperfusion. Amounts of the test solution equal to 10% of the dog's blood volume were administered intra-atrially to the animals just before release of the LAD clamp. Six dogs received Ringer's lactate, 7 were given 600,000 IU of SOD, 13 received 6% HES-Pz, and 9 were given SOD and HES-Pz. The ratio of infarct to area at risk was 20 +/- 3% in the control dogs receiving Ringer's lactate, 16 +/- 4% in animals receiving SOD alone (p = NS), 6 +/- 3% in dogs receiving HES-Pz alone (p < 0.05), and 8 +/- 3% in dogs given a combination of SOD and HES-Pz (p < 0.05). HES-Pz alone and with SOD significantly reduced reperfusion injury, although addition of SOD to HES-Pz did not have an additive effect. Appropriate-sized macromolecules may act by reducing ischemia-induced microvascular permeability.