Objective To evaluate prenatal corticosteroid use in women experiencing spontaneous preterm labor and preterm delivery. Methods The present cross‐sectional multicenter study analyzed interview data from patients attending 20 hospitals in Brazil owing to preterm delivery between April 1, 2011 and July 30, 2012. Patients were stratified based on preterm delivery occurring before 34 weeks or at 34–36+6 weeks of pregnancy, and the frequency of prenatal corticosteroid use at admission was compared. Prenatal corticosteroid use, sociodemographic data, obstetric characteristics, and neonatal outcomes were examined. Results There were 1455 preterm deliveries included in the present study; 527 (36.2%) occurred before 34 weeks of pregnancy and prenatal corticosteroids were used in 285 (54.1%) of these pregnancies. Among neonates delivered at 32–33+6 weeks, prenatal corticosteroid use was associated with lower pneumonia (P=0.026) and mortality (P=0.029) rates. Among neonates delivered at 34–36+6 weeks, prenatal corticosteroid use was associated with longer neonatal hospital admission (P<0.001), and an increased incidence of 5‐minute Apgar scores below 7 (P=0.010), endotracheal intubation (P=0.042), surfactant use (P=0.006), neonatal morbidities (P=0.048), respiratory distress (P=0.048), and intraventricular hemorrhage (P=0.023). Conclusion Preterm labor and late preterm delivery were associated with worse neonatal outcomes following prenatal corticosteroids. This could reflect a sub‐optimal interval between administration and delivery. Prenatal corticosteroids were associated with worse neonatal outcomes following late preterm delivery; this could reflect a sub‐optimal interval between final corticosteroid dose and delivery.