To investigate trends in birth prevalence of cerebral palsy (CP) overall and by gestational age, and examine the distribution of motor type, spastic topography, and severity using Australian CP Register data from 1995 to 2009. Prenatal and perinatal CP data were collated from state/territory CP registers. Birth prevalence estimates per 1000 live births and per 1000 neonatal survivors (NNS) were calculated in five epochs. Data from three state registers with population-level ascertainment were used to investigate birth prevalence trends by gestational age using Poisson regression. Distribution of motor type, spastic topography, and moderate to severe disability (IQ≤50 and/or Gross Motor Function Classification System levels III-V) were evaluated within birthweight categories. Birth prevalence of CP varied across population-level states but within each state declined significantly over time (p<0.05). Birth prevalence per 1000 neonatal survivors declined amongst children born before 28 weeks (South Australia, Victoria p<0.001) and those born at or after 37 weeks (Victoria p<0.001, Western Australia p<0.002). Across Australia the percentage of children with bilateral spastic CP declined amongst those born less than 1000g. The percentage of children with moderate to severe disability decreased (48%-34%, p<0.001). Birth prevalence of CP declined. Encouragingly, the percentage of children with CP whose disability was moderate to severe also decreased. Birth prevalence of cerebral palsy (CP) differed but declined across Australian states (1995-2009). Australian CP birth prevalence declined significantly amongst children born before 28 weeks and those born at or after 37 weeks. The percentage of children with moderate to severe disability decreased.