Abstract
Eighty one very low birthweight survivors with cerebral palsy were matched with controls by sex, gestational age, and place of birth. Using discriminant analysis, the perinatal profiles for infants with cerebral palsy and their controls were shown to differ significantly. When infants with various types of cerebral palsy were analysed with their controls the discriminating variables differed. Diplegic infants could be differentiated from controls on antenatal variables alone, but significant discrimination of hemiplegic and quadriplegic infants required the addition of postnatal variables. Cranial ultrasound appearances differed appreciably between types of cerebral palsy. Future studies should differentiate between types of cerebral palsy and include ultrasound data. Cerebral palsy in very low birthweight infants is unlikely to prove a useful outcome indicator for neonatal intensive care.