Objective: To assess in a single cohort whether annual weight and waist circumference (WC) change has varied over time. Design: Longitudinal cohort study with three surveys (1) 1999/2000; (2) 2004/2005 and (3) 2011/2012. Generalised linear mixed models with random effects were used to compare annualised weight and WC change between surveys 1 and 2 (period 1) with that between surveys 2 and 3 (period 2). Models were adjusted for age to analyse changes with time rather than age. Models were additionally adjusted for sex, education status, area-level socioeconomic disadvantage, ethnicity, body mass index, diabetes status and smoking status. Setting: The Australian Diabetes, Obesity and Lifestyle study (AusDiab).a population-based, stratified-cluster survey of 11247 adults aged ≥25 years. Participants: 3351 Australian adults who attended each of three surveys and had complete measures of weight, WC and covariates. Primary outcome measures: Weight and WC were measured at each survey. Change in weight and WC was annualised for comparison between the two periods. Results: Mean weight and WC increased in both periods (0.34 kg/year, 0.43 cm/year period 1; 0.13 kg/year, 0.46 cm/year period 2). Annualised weight gain in period 2 was 0.11 kg/year (95% CI 0.06 to 0.15) less than period 1. Lesser annual weight gain between the two periods was not seen for those with greatest arealevel socioeconomic disadvantage, or in men over the age of 55. In contrast, the annualised WC increase in period 2 was greater than period 1 (0.07 cm/year, 95% CI 0.01 to 0.12). The increase was greatest in men aged 55+ years and those with a greater area-level socioeconomic disadvantage. Conclusions: Between 2004/2005 and 2011/2012, Australian adults in a national study continued to gain weight, but more slowly than 1999/2000.2004/2005. While weight gain may be slowing, this was not observed for older men or those in more disadvantaged groups, and the same cannot be said for WC.