We enrolled 280 intubated babies with birth weights of less than 1,751 g in a double-blind randomized prospective clinical trial to evaluate whether phenobarbital influences the likelihood of developing subependymal-intraventricular-intraparenchymal hemorrhage. Phenobarbital was associated with an increased risk of developing any subependymal-intraventricular-intraparenchymal hemorrhage and was not associated with a diminished risk of either severe hemorrhage or germinal matrix hemorrhage. This increased risk was apparent even after we considered the influence of phenobarbital levels, timing of phenobarbital administrations, institutional differences, quality of ultrasound scans, gestational age- and birth weight-specific effects, ascertainment bias, and other possible confounders of phenobarbital administration.