During renal ischemia, ATP is degraded to hypoxanthine. When xanthine oxidase converts hypoxanthine to xanthine in the presence of molecular oxygen, superoxide radical (O2-) is generated. We studied the role of O2- and its reduction product OH · in mediating renal injury after ischemia. Male Sprague-Dawley rats underwent right nephrectomy followed by 60 min of occlusion of the left renal artery. The O2- scavenger superoxide dismutase (SOD) was given 8 min before clamping and before release of the renal artery clamp. Control rats received 5% dextrose instead. Plasma creatinine was lower in SOD treated rats: 1.5, 1.0, and 0.8 mg/dl vs. 2.5, 2.5, and 2.1 mg/dl at 24, 48, and 72 h postischemia. 24 h after ischemia inulin clearance was higher in SOD treated rats than in controls (399 vs. 185 μl/min). Renal blood flow, measured after ischemia plus 15 min of reflow, was also greater in SOD treated than in control rats. Furthermore, tubular injury, judged histologically in perfusion fixed spedimens, was less in SOD treated rats. Rats given SOD inactivated by prior incubation with diethyldithiocarbamate had plasma creatinine values no different from those of control rats. The OH · scavenger dimethylthiourea (DMTU) was given before renal artery occlusion. DMTU treated rats had lower plasma creatinine than did controls: 1.7, 1.7, and 1.3 mg/dl vs. 3.2, 2.2, and 2.4 mg/dl at 24, 48, and 72 h postischemia. Neither SOD nor DMTU caused an increase in renal blood flow, urine flow rate, or solute excretion in normal rats. The xanthine oxidase inhibitor allopurinol was given before ischemia to prevent the generation of oxygen free radicals. Plasma creatinine was lower in allopurinol treated rats: 2.7, 2.2, and 1.4 mg/dl vs. 3.6, 3.5, and 2.3 mg/dl at 24, 48, and 72 h postischemia. Catalase treatment did not protect against renal ischemia, perhaps because its large size limits glomerular filtration and access to the tubular lumen. Superoxide-mediated lipid peroxidation was studied after renal ischemia. 60 min of ischemia did not increase the renal content of the lipid peroxide malondialdehyde, whereas ischemia plus 15 min reflow resulted in a large increase in kidney lipid peroxides. Treatment with SOD before renal ischemia prevented the reflow-induced increase in lipid peroxidation in renal cortical mitochondria but not in crude cortical homogenates. In summary, the oxygen free radical scavengers SOD and DMTU, and allopurinol, which inhibits free radical generation, protected renal function after ischemia. Reperfusion after ischemia resulted in lipid peroxidation; SOD decreased lipid peroxidation in cortical mitochondria after renal ischemia and reflow. We conclude that restoration of oxygen supply to ischemic kidney results in the production of oxygen free radicals, which causes renal injury by lipid peroxidation.