In a prospective, randomized, double-blind, multicentre trial the effect of antenatal treatment with betamethasone phosphate was compared with placebo in the prevention of the respiratory distress syndrome (RDS) in preterm infants. The dose of betamethasone was 4 mg every 8 h for six doses, unless delivery occurred. The 251 women who were enrolled gave birth to 262 liveborn infants, 130 in the beta-methasone and 132 in the placebo group; the two groups were evenly matched in most respects. The diagnosis of RDS in the newborn was confirmed by two independent assessors. Seven of the 130 infants in the betamethasone group and 16 of the 132 in the placebo group developed RDS. In infants whose mothers had received at least three injections, RDS was also less frequent in the steroid group than in the placebo group (3/104 and 10/104 respectively; P less than 0.05). There was a significant reduction of RDS in those born between 24 h and 6 days after entry into the trial (0/30 and 8/45 respectively; P less than 0.05). The largest difference in frequency of RDS occurred in the subgroup of infants born before 34 weeks gestation, within 8 days of trial entry, and whose mothers had received at least three injections (0/27 steroid group and 7/32 placebo group; P = 0.03), and there were also significantly fewer neonatal deaths (2/27 and 13/32, respectively; P less than 0.01) in this subgroup. Betamethasone did not provoke earlier delivery. Premature rupture of the membranes and maternal hypertension did not seem to contraindicate the use of steroids: there was no increase in maternal or neonatal sepsis nor in stillbirth in hypertensive pregnancies in the steroid group. Neonatal jaundice was significantly less frequent in the steroid (55/129) than in the placebo group (81/127; P less than 0.01) but not in the subgroups born before 34 completed weeks gestation.