BACKGROUND: Popular media, health experts and researchers talk about a paediatric 'obesity epidemic' with exponentially increasing rates of obesity and overweight. However, some recent reports suggest that prevalence may have plateaued. This study examined trends in the prevalence of Australian childhood overweight and obesity since 1985. Specifically, it aimed to determine whether there have been (a) overall increases in average body mass index (BMI), (b) differential patterns of change within age groups and (c) increases in BMI within each weight-status category. METHOD: Forty-one Australian studies of childhood weight status conducted between 1985 and 2008 were reviewed. The studies included data on 264 905 Australians aged 2-18 years, with raw data being available on 70 758 children (27%). Children were classified as overweight or obese based on BMI using the criteria of Cole et al. (BMJ, 2000). The prevalence estimates were adjusted for age and sex, and plotted against measurement year using Lowess plots and two-linear-segment models. Where raw data were available, BMI z-scores (UK 1990 standard) were plotted against measurement year for all children and children in various age groups. Lowess plots and two-linear-segment models were used to assess secular trends in BMI z-scores pre- and post-1996 within age, gender and weight-status categories. RESULTS: There has been a plateau, or only slight increase, in the percentage of boys and girls classified as overweight or obese, with almost no change over the last 10 years. In boys and girls, prevalence rates have settled around 21-25% for overweight and obesity together, and 5-6% for obesity alone. Similar trends were found for BMI z-scores. These patterns were fairly consistent across the age span. Within each weight-status category, average BMI has not increased. CONCLUSIONS: Although levels of Australian paediatric overweight remain high, the prevalence of overweight and obesity seems to have flattened and has not followed the anticipated exponential trajectory.