QUESTION: What is the free-living physical activity of community-dwelling people with stroke compared with that of age-matched healthy controls? DESIGN: A cross-sectional observational study. PARTICIPANTS: 42 people with stroke and 21 age-matched healthy controls aged 52 to 87 years living in Sydney, Australia. OUTCOME MEASURES: Free-living physical activity was measured using the Intelligent Device for Energy Expenditure and Activity (IDEEA) and reported as duration (time on feet in min) and frequency (activity counts). RESULTS: People with stroke spent 79 (95% CI 20 to 138) fewer min on their feet and performed 5308 (95% CI 3171 to 7445) fewer activity counts than healthy controls. The observation period of the free-living physical activity of stroke survivors was significantly less than that of the healthy controls. Data adjusted to a fixed observation period (12 hr) showed no relative difference in time on feet between the groups (mean difference 36 min, 95% CI -27 to 99) but that people after stroke still had relatively fewer activity counts than healthy controls (mean difference 4062 counts, 95% CI 1787 to 6337). CONCLUSIONS: The reduction in physical activity after stroke is not primarily because of a decrease in the time spent being active but rather a decrease in frequency of activity during that time. Future research into physical activity after stroke needs to consider energy expenditure because stroke survivors exhibit a reduced frequency of physical activity due to the nature of their impairments.