We compared the neurodevelopmental outcome of extremely premature, surfactant-deficient infants who received either prophylactic surfactant at birth, "rescue" surfactant after the clinical diagnosis of respiratory distress syndrome was established, or placebo. Infants studied were participants in a randomized, bicenter (San Diego, Calif., and Helsinki, Finland), controlled trial of human surfactant therapy. One hundred fifty infants (prophylaxis group, 63 infants; rescue group, 57; placebo group, 30) were prospectively enrolled at 38 weeks of gestational age. There were no neonatal intergroup differences in the incidence or severity of sonographic central nervous system abnormality or retinopathy. One hundred forty-five infants were alive at 1 year of adjusted age, at which time growth, neurosensory, and neurologic outcome were similar in all three treatment groups at both centers. Cerebral palsy occurred in 20% overall. Five infants (3.5%) were functionally blind. However, infants treated at birth had lower mean mental and motor scores on the Bayley Scales of Infant Development compared with those of infants rescued with surfactant after the onset of respiratory distress syndrome (Mental Development Index: 78 vs 96, p = 0.02; Psychomotor Development Index: 73 vs 87, p = 0.04). Chronic lung disease occurred more frequently in the prophylactically treated group and contributed to the subjects' neurologic and developmental morbidity. Because prophylactic surfactant treatment offered no neurodevelopmental advantage and may contribute to poorer outcome, we currently recommend early surfactant replacement only for those infants who have postnatal evidence of respiratory distress syndrome.