We enrolled 61 neonates of 600 to 1250 gm birth weight with evidence of low-grade intraventricular hemorrhage at 6 to 11 hours of age in a prospective, randomized, placebo-controlled trial to test the hypothesis that indomethacin (0.1 mg/kg given intravenously at 6 to 12 postnatal hours and every 24 hours for two more doses) would prevent extension of intraventricular hemorrhage. Twenty-seven infants were assigned to receive indomethacin; 34 infants received saline placebo. There were no significant differences between the two groups in birth weight, gestational age, sex, Apgar scores, percentage of infants treated with surfactant, or distribution of hemorrhages at the time of the first cranial sonogram (echo-encephalogram). Within the first 5 days, 9 of 27 indomethacin-treated and 12 of 34 saline solution-treated infants had extension of their initial intraventricular hemorrhage (p = 1.00). Four indomethacin-treated and three saline solution-treated infants had parenchymal extension of the hemorrhage. Indomethacin was associated with closure of a patent ductus arteriosus by the fifth day of life (p = 0.003). There were no differences in adverse events attributed to indomethacin. We conclude that in very low birth weight infants with low grade intraventricular hemorrhage within the first 6 postnatal hours, prophylactic indomethacin therapy promotes closure of the patent ductus arteriosus and is not associated with adverse events, but does not affect the cascade of events leading to parenchymal involvement of intracranial hemorrhage.