A multicenter, randomized, double-blind, controlled trial compared three beractant (Survanta) administration procedures in the treatment of neonatal respiratory distress syndrome. Infants weighing > or = 600 gm with respiratory distress syndrome who required assisted ventilation were treated within 8 hours of birth with beractant administered intratracheally. Procedure A required administration in two fractional doses after removal of the infant from the ventilator. Procedure B required administration in two fractional doses through a neonatal suction valve and did not require removal of the infant from the ventilator, and procedure C required administration in four fractional doses during removal from the ventilator. Procedure C is the method used in all previous beractant studies. Of the 299 infants enrolled, 103 were randomly assigned to procedure A, 100 to procedure B, and 96 to procedure C. The results indicate no significant differences among the treatment groups in the clinical outcome measures of fractional inspired oxygen, mean airway pressure, and arterial-alveolar ratio of partial pressure of oxygen at 72 hours of life, or in the incidences of air leaks, pulmonary interstitial emphysema, or death through 72 hours of life. There were no significant differences in the lowest heart rates recorded during administration of doses, but there was less oxygen desaturation during administration of dose 1 with procedure B than with procedure A (p = 0.001), and more reflux of beractant after procedure B than after procedure C (p = 0.007). We conclude that the three procedures are equally effective and can be used to administer beractant safely and effectively. Procedure B has the advantage of allowing administration without interrupting mechanical ventilation.