Background The Olympic Games’ (OG) organisers typically hope that a diverse range of health legacies, including increases in physical activity and sport participation will result from their hosting of the OG. Despite these aspirations, the effects of the Olympics on physical activity levels remain to be demonstrated in large-scale population studies.

Methods This study examined the short-term impact of the Sydney 2000 OG, using serial cross-sectional population physical activity surveys of Australian adults in November 1999 and November 2000. Random sample surveys of adults asked about physical activity participation, intention to be more active, and in 2000, response to the OG.

Results There were no significant effects of the Olympics on physical activity participation among adult Australians, measured 6 weeks after the end of the Games. Total minutes of leisure-time physical activity did not change significantly between 1999 and 2000 (295–303 min/week), and the proportion reaching the recommended levels of 150 min/week did not change (56.6% and 56.8%, respectively, in 1999 and 2000). The intention to be active in the next month increased after the Games (adjusted OR 1.12, 95% CI 1.01 to 1.24), but was not associated with physical activity behaviour change.

Conclusions The legacy of the OG may be apparent through new infrastructure and other urban improvements, but evidence of their influence on physical activity levels remains elusive. Without multiyear integrated and well-funded programmes to promote physical activity, the Olympic legacy of a more active community may remain more rhetoric than reality.