OBJECTIVE: Our purpose was to determine the efficacy of maternal corticosteroid therapy between 26 and 31 weeks' gestation. STUDY DESIGN: The data in this study were derived from 32,658 women who participated in the March of Dimes-sponsored multicenter prematurity prevention program. Of the 432 women who were delivered at 26 to 31 weeks, 67 received betamethasone before delivery and 365 did not. The frequency and relative risks of adverse outcomes, including respiratory distress syndrome, intraventricular hemorrhage, necrotizing enterocolitis, and neonatal death were compared for each of two gestational age periods by means of univariate and multivariate techniques. RESULTS: When betamethasone was administered > or = 2 days before delivery (n = 45), there was a lower incidence of respiratory distress syndrome in both the 26 to 28 week group (53.9% vs 86.5%, p = 0.008) and the 29 to 31 week group (25.0% vs 59.1%, p = 0.0003). The rate of intraventricular hemorrhage was less in the betamethasone group at 26 to 28 weeks (15.4% vs 32.3%, p = 0.17), but the difference reached statistical significance only at 29 to 31 weeks (3.1% vs 16.5%, p = 0.029). Neonatal death occurred significantly less often in infants who were delivered at 26 to 28 weeks when their mothers received betamethasone compared with infants of the same gestational age whose mothers did not receive betamethasone treatment (0% vs 34.6%, p = 0.01). In a regression analysis of infants born between 26 and 31 weeks in which birth weight, gestational age, race, infant sex, and tocolytic use were controlled, the odds ratio for respiratory distress syndrome associated with betamethasone use was 0.20 (0.10, 0.42), for intraventricular hemorrhage 0.26 (0.08, 0.90), and for neonatal death 0.14 (0.02, 1.09). Insufficient numbers of women were given betamethasone before 26 weeks for analysis. CONCLUSION: Betamethasone appears to significantly reduce neonatal death and the morbidity between 26 and 31 weeks' gestation.