Patent ductus arteriosus (PDA) is believed to be a contributing factor in the etiopathogenesis of bronchopulmonary dysplasia (BPD). We studied the effects of early dexamethasone therapy on persistent ductal patency and the role of PDA in the etiopathogenesis of BPD during the course of a randomized double-blind trial of dexamethasone to prevent BPD. Infants, who weighed between 700 and 999 g, had severe RDS, and had been given surfactant, were randomized to receive a 12-day course of dexamethasone (n = 13) or placebo (n = 17) starting within the first 12 hours of postnatal life. The diagnosis of PDA was made clinically and was confirmed by cardiac ultrasound. The incidence of clinically significant ductus in infants who weighed less than 1000 g was 23% in the dexamethasone-treated group, as compared with 59% in infants who were given placebo. This difference was marginally significant, p = 0.05, odds ratio 0.21, 95% confidence interval 0.04-1.05. None of the infants in the dexamethasone group had recurrence of PDA after indomethacin therapy as compared with three infants in the placebo group. Dexamethasone significantly reduced the number of days infants required ventilator and supplemental oxygen as compared with infants who received placebo. Dexamethasone, as compared with placebo, also reduced the incidence of BPD, p = 0.025, odds ratio 0.08, 95% confidence interval 0.01-0.58. Dexamethasone may reduce the incidence of PDA in premature infants who weigh less than 1000 g at birth and thereby reduce the incidence of BPD.