BACKGROUND: To develop a model for prediction of severe intracranial hemorrhage (ICH) or death based on variables from the first 12 h of age and to compare mortality and morbidities with and without exposure to early indomethacin. METHODS: This retrospective cohort study included extreme preterm (220/7−266/7 weeks) infants born at National Institute of Child Health and Human Development Neonatal Research Network sites. Primary outcome was a composite of severe ICH and/or death. RESULTS: Of 4624 infants, 1827 received early indomethacin. Lower gestation, lack of antenatal steroids exposure, lower 1-min Apgar, male sex, and receipt of epinephrine were associated with severe ICH or death. Early indomethacin was associated with a lower risk of patent ductus arteriosus, bronchopulmonary dysplasia, and higher risk of spontaneous intestinal perforation. CONCLUSIONS: A model for early prediction of severe ICH/death was developed and validated. Early indomethacin was associated with a lower risk of patent ductus arteriosus and bronchopulmonary dysplasia and a higher risk of spontaneous intestinal perforation.