The effectiveness of bovine superoxide dismutase (SOD) in the prevention of bronchopulmonary dysplasia was evaluated in a prospective double-blind controlled study in 45 neonates (mean gestational age 28.7 weeks, mean weight 1154 gm) with severe respiratory distress syndrome. All were ventilator dependent with FiO2 greater than or equal to 0.7 at 24 hours of age. Either bovine SOD (0.25 mg/kg) or saline solution was administered subcutaneously every 12 hours according to random selection until patients could be maintained in room air without ventilatory or continuous positive airway pressure (CPAP) support. SOD levels were detected in all patients given treatment. Mean peak values at 4 hours after dose ranged from 0.15 micrograms/ml (dose 1) to 0.45 micrograms/ml (dose 10). The drug was well tolerated, and no side effects were detected. Among the 31 survivors (SOD 14, placebo 17) radiologic evidence of BPD was significantly less in patients given SOD (3/14 vs 12/17, P = 0.008). Clinical signs of bronchopulmonary dysplasia (wheezing, pneumonia) were less in patients given SOD (3/14 vs 11/17, P = 0.019). Patients given SOD required fewer days of CPAP (P less than 0.003). There were no differences in days of O2 therapy, intermittent positive pressure breathing, or incidence and severity of patent ductus arteriosus or intraventricular hemorrhage. This preliminary study suggests that SOD may be helpful in reducing the severity of bronchopulmonary dysplasia in infants with respiratory distress syndrome.