Impaired psychomotor development, often anteceded by major intraventricular hemorrhage or periventricular leukomalacia, constitutes the most important long-term morbidity of very preterm infants. We reviewed randomized controlled trials aimed at reducing the incidence of brain damage, as detected by ultrasound, or neurodevelopmental impairment during follow-up of preterm infants. Preliminary reports of reduced rates of intraventricular hemorrhage obtained with administration of fresh frozen plasma, ethamsylate, phenobarbitone, or morphine have not been confirmed in subsequent larger trials. Early administration of indomethacin may reduce intraventricular hemorrhage without affecting long-term outcome. Pancuronium, inositol, and vitamin E decreased intraventricular hemorrhage rates but later psychomotor development was not examined. Thyroxin supplementation failed to improve neurodevelopmental outcome while protein enrichment of formula and individualized developmental care appear to be beneficial. The largest reductions in cerebral palsy and neurodevelopmental impairment were achieved by avoidance of postnatal steroids. This finding emphasizes the need to include these late endpoints in any randomized trial involving preterm infants.