The antecedents of a population of children with neurodevelopmental disability were examined in a sibling-control study: siblings were used to control for the effect of the family environment on development. Of 294 children with neurodevelopmental disability, 228 who had normally developing siblings were matched with the sibling of nearest ordinal position in the family. There was a significantly higher incidence of prenatal complications among disabled children than among their matched siblings, even after correction for gestational age at birth. Prenatal complications and delayed onset of regular respiration were predictive of disability. These data confirm the authors' observation that the relation between prenatal complications and neurodevelopmental disability is independent of social class and gestational age at birth.