Many reports suggest that physically active women have a somewhat lower breast cancer incidence than physically inactive women. We hypothesized that indices of physical activity are associated inversely with breast cancer incidence after adjustment for confounders. The sample comprised 7994 women, aged 45-64, who participated in the Atherosclerosis Risk in Communities (ARIC) Study. Baseline physical activity was assessed by the Baecke questionnaire. Over an average follow-up of 13.1 yrs, 342 incident breast cancer cases were ascertained. After adjustment for age, race, study center, age at first live birth, age at menopause, and family history of breast cancer in a first-degree relative, there was no statistically significant association of breast cancer incidence with baseline physical activity levels for leisure, sport or work indices. Compared with the lowest quartile of physical activity, women in the highest quartile had a multivariate-adjusted hazard ratio (HR) of 1.00 (95% confidence interval (CI)=0.64-1.54) for the leisure index, 1.31 (95% CI=0.87-1.96) for the sport index and 0.87 (95% CI=0.61-1.24) for the work index. Our findings do not corroborate the majority of previous reports, which implicated physical inactivity as a risk factor for breast cancer.