We performed a multicenter prospective randomized controlled trial to determine the efficacy and safety of the surfactant preparation, Survanta (Abbott Laboratories, Chicago, USA), for 750-1750 g infants with idiopathic respiratory distress syndrome, (IRDS) receiving assisted ventilation with 40% or more oxygen. One hundred and six eligible infants from the eight participating centers were randomly assigned between March 1986 and June 1987 to receive either surfactant (100 mg phospholipid/kg, 4 ml/kg) or air (4 ml/kg) administered into the trachea within 8 h of birth (median time of treatment 6.2 h, range 3.2-9.1 h). The study was stopped before enrollment was completed at the request of the United States Food and Drug Administration when significant differences were observed in incidence of periventricular-intraventricular hemorrhage (PIH), between the surfactant treated and control infants. Surfactant treated infants had larger average increases in the arterial-alveolar oxygen ratio, (a/A ratio) (P less than 0.0001), and larger average decreases in FiO2 (P less than 0.0001) and mean airway pressure, (MAP) (P less than 0.017) than controls over the 48 h following treatment. The magnitude of the differences between the surfactant and control groups were 0.19 (SE = 0.03) for a/A ratio, -0.28 (SE = 0.04) for FiO2 and -1.7 cm H2O (SE = 0.70) for MAP. The clinical status on days 7 and 28 after treatment was classified using four predefined ordered categories: (1) no respiratory support; (2) supplemental O2 with or without continuous positive airway pressure (CPAP); (3) intermittent mandatory ventilation; and (4) death. There were no statistically significant differences in the status categories on days 7 or 28 between surfactant and control infants.(ABSTRACT TRUNCATED AT 250 WORDS)