tently higher proportions of CP in the Indigenous population. The three jurisdictions in Australia with the largest Indigenous populations are WA, Northern Territory (NT) and Queensland (Qld). As CP registers in NT and Qld are now considered to have a sufficient level of ascertainment, data from the three registers have been combined to compare the occurrence of CP in their Indigenous and non-Indigenous populations. Aim: To compare proportions of live births subsequently described as having CP, the distributions of associated impairments and causes of postneonatal (PNN) CP between Indigenous and non-Indigenous populations in Australia. Method: Data from statutory birth records and CP registers for 1996–2005 birth cohorts in WA, NT and Qld were stratified by Indigenous status and whether the CP was acquired pre/perinatally or postneonatally. Relative risks associated with Indigenous status were estimated and the distributions of causes of PNN CP compared. Results: Indigenous births had a relative risk of 4.9 (95% confidence interval 3.0, 7.9) for PNN CP but only of 1.42 (95% CI 1.2, 1.7) for pre/perinatal CP. Almost half of PNN CP in Indigenous infants resulted from infection, whereas for nonIndigenous infants the most frequent cause was cerebrovascular accident. The impairments of Indigenous CP and of PNN CP tended to be more numerous and more severe. Conclusion: Indigenous children are at significantly greater risk of CP, particularly PNN CP. The predominant cause of PNN CP in non-Indigenous children has shifted to cerebrovascular accident over time; however, infections followed by head injury are still the most frequent causes in Indigenous infants.