Data on all children with spastic cerebral palsy (N = 183) and on a matched group of control children (N = 549) born in Western Australia between 1975 and 1980 were compared to investigate the relationship between birth asphyxia and spastic cerebral palsy. Information on perinatal events for both the children with cerebral palsy and the control subjects was collected by means of epidemiologic methods to reduce bias. An association between clinically observed perinatal signs of birth asphyxia and spastic cerebral palsy was found (relative risk 2.84; 95% confidence interval 1.85 to 4.37). The population-attributable risk proportion was 14.1%. The likelihood of birth asphyxia's causing perinatal brain damage was assessed by two independent observers using defined criteria. It was estimated that in only about 8% (15/183) of all the children with spastic cerebral palsy was intrapartum asphyxia the possible cause of their brain damage. The contribution of intrapartum events and obstetric mismanagement to overall cerebral palsy rates is probably less than was previously thought.