To test whether early postnatal (0 to 3 hours) intravenous administration of ibuprofen will prevent patent ductus arteriosus (PDA) in preterm neonates. Prospective sequential controlled trial with three treatment arms. Level 3 perinatal-neonatal intensive care nursery. Thirty-four premature newborn infants born from February to August 1993 with a mean birth weight of 913 g (range, 565 to 1460 g) and gestational age of 26.9 weeks (range, 22.4 to 31.0). Infants were consecutively assigned within 3 hours of age to treatment with either one dose of ibuprofen lysine (10 mg/kg intravenously) followed by 5 mg/kg per dose intravenously at 24 and 48 hours of age (n = 12), one dose of ibuprofen lysine (10 mg/kg intravenously; n = 11), or saline (n = 11). Primary outcome variable was the presence of ductus arteriosus by echocardiography and clinical assessments at 3, 7, and 21 days of life. Secondary outcome variables included presence of intraventricular hemorrhage, renal function, ventilatory and oxygen needs, hematologic changes, gastrointestinal function, time to full enteral feeding, duration of hospitalization, and age at discharge. The three groups of patients were comparable in birth weight, gestational age, antenatal administration of betamethasone, and other perinatal characteristics. Ibuprofen treatment significantly reduced plasma levels of prostaglandins, and the levels remained low for 72 hours in newborns who received three doses of the drug. The incidence of PDA and other variables did not differ between patients who received a single dose of ibuprofen and those given saline. However, compared with the saline-treated newborns, babies who received three doses of ibuprofen had no PDA (0/12 vs 7/11 for saline; P < .02), had lower daily mean airway pressures (mean +/- SD, 5.2 +/- 1.1 cm H2O vs 8.3 +/- 2.8 cm H2O for saline; P < .02) and better oxygenation index (2.6 +/- 0.6 vs 4.7 +/- 1.8 for saline; P < .02) at the end of the first week of life, and required fewer days of ventilation (25 +/- 14 days vs 44 +/- 26 days for saline; P < .03). Babies given three doses of ibuprofen tended to tolerate full oral feedings earlier (35 +/- 19 days vs 56 +/- 34 days for saline; P = .09), had shorter duration of hospitalization (71.2 +/- 22.6 days vs 127.3 +/- 74.7 days for saline; P < .05), and were discharged to home at an earlier postconceptional age (37.8 +/- 2.0 weeks vs 44.8 +/- 9.8 weeks for saline; P < .05). ibuprofen treatment in this phase I trial was not associated with any apparent early neurological, intestinal, renal, hepatic, or hematologic complications. Administration of three doses of ibuprofen within 3 hours after birth in preterm neonates reduced the incidence of PDA without causing notable early adverse drug reactions in this phase I trial. Early closure of the ductus arteriosus was also associated with better respiratory outcome and earlier discharge from the hospital.