The effectiveness of human superoxide dismutase (hSOD) in the prevention of reperfusion injury was evaluated in a rabbit ileal loop model. Weanling white New Zealand rabbits, 6 weeks of age and weighing 500 to 1,000 g, were used. Intraluminal administration of SOD (5 mg/kg) was studied in 12 animals with each animal serving as its own control. In an additional 12 animals, parenteral SOD in a dose of 5 mg/kg in seven animals and 10 mg/kg in five animals was evaluated, while five additional control animals received parenteral saline. The effect of reperfusion injury was evaluated in each bowel loop by interruption of blood supply for five minutes, followed by reperfusion. Blood was drawn at 0, 16, 20, 24 hours in the parenteral group for measurement of hSOD levels by radioimmunoassay. The loops were studied pathologically for extent of mucosal damage. In the intraluminal group, nine of 12 loops without SOD v three of 12 loops with SOD showed necrosis when rendered ischemic (P = .0196). In the parenteral group 22 of 24 loops were normal when pretreated with SOD and subjected to ischemia v five of ten when no SOD was given (P = .0139). In the parenteral group, mean baseline level of hSOD was 0.42 +/- 0.26 micrograms/mL. Levels peaked at 16 hours (3.64 +/- 1.75 micrograms/mL) and progressively decreased at 20 hours (2.85 +/- 1.34 micrograms/mL) and 24 hours (1.82 +/- 1.15 micrograms/mL). This preliminary animal study suggests that hSOD may be an effective method for the prevention of postischemic bowel injury, adding to the literature on the protective effects of SOD in various models of intestinal ischemia.