We undertook a randomized, controlled trial to determine whether human surfactant administered endotracheally at birth to very premature infants (gestational age, 24 to 29 weeks) would prevent the respiratory distress syndrome or reduce its severity. Thirty-one treated infants (birth weight, 938 +/- 286 g) were compared in a blinded fashion with 29 control infants (birth weight, 964 +/- 174 g). The lecithin/sphingomyelin ratio was less than 2 in all infants, and phosphatidylglycerol was not present in amniotic fluid or tracheal fluids at birth, indicating a deficiency of surfactant in the lungs. The principal dependent variables were neonatal death, the incidence of bronchopulmonary dysplasia, and the infant's requirement for respiratory support (and its complications). The surfactant-treated group had significantly fewer deaths than the control group (16 percent vs. 52 percent, P less than 0.001), fewer cases of bronchopulmonary dysplasia (16 percent vs. 31 percent), and significantly fewer cases of pulmonary interstitial emphysema (P less than 0.001) and pneumothorax (P less than 0.02). Prophylactic treatment with human surfactant also substantially reduced the period of neonatal intensive care. We conclude that treatment with human surfactant offers promise for improving the survival of very premature infants with a surfactant deficiency and for reducing the pulmonary sequelae of the respiratory distress syndrome.