Superoxide radicals formed during reperfusion of ischemic tissues have been identified as a key mediator in the microvascular manifestations of postischemic tissue damage. This understanding is based on studies in laboratory animals in which high doses of superoxide dismutase (SOD; 2.0-25.0 mg/kg body wt iv) were found to inhibit postischemic leukocyte adhesion and the leakage of fluid and macromolecules. Using a dorsal skinfold chamber model in hamsters, we demonstrate now that protection from reperfusion-induced leukocyte adhesion to venular endothelium after 4 h of ischemia to striated muscle can be attained by pretreatment of the animals with a significantly lower dose of exogenous CuZn-SOD (0.25 mg/kg body wt) or with heparin (2,000 IU/kg body wt), which induces a comparable increase in SOD plasma activity through the release of endogenous extracellular SOD from endothelial cell binding sites. This protective effect was maintained until 24 h after reperfusion. In contrast, CuZn-SOD or heparin failed to attenuate the postischemic shutdown of nutritional capillary perfusion, a phenomenon that is due to ischemia-induced endothelial cell swelling, rather than due to reperfusion-associated events, and hence is not susceptible to strategies directed against oxygen radicals generated during the reperfusion phase. The results of this study 1) imply that postischemic leukocyte/endothelium interaction can be attenuated by a low and clinically more relevant dose of SOD, and 2) caveat the administration of heparin in laboratory animals (i.e., to keep catheters patent) in studies of experimental ischemia/reperfusion injury or other oxygen radical-dependent pathomechanisms.