In 1994, the National Institutes of Health Consensus Development Conference on Antenatal Steroids recommended corticosteroids between 24 and 30-32 weeks' gestation in pregnancies complicated by preterm premature rupture of membranes (PROM). Since the Consensus Conference, the use of antenatal corticosteroids has increased to approximately 60% of potential treatment candidates. Some of the remaining 40% of pregnant candidates may go untreated because of concern that corticosteroids could increase the risk of neonatal infection. Using decision-analysis techniques, we compared the potential benefit of antenatal corticosteroids in reducing the incidence of severe intraventricular hemorrhage with the potential risk of increasing the rate of neonatal sepsis. Our analysis indicates that the benefit of a small decrease in severe intraventricular hemorrhage outweighs the potential harm of a large increase in the rate of neonatal sepsis. Therefore, we support the Consensus Conference panel's recommendation that antenatal corticosteroids be used in pregnancies complicated by preterm PROM.