The association between recreational physical activity and breast cancer risk was examined in a population-based case-control study in Adelaide, Australia between 1982 and 1984. There were 451 incident, histologically confirmed cases of breast cancer, identified through the South Australian Central Cancer Registry, which were each age-matched to one control selected at random from the electoral register. These women, aged 20-74 years at diagnosis, reported their level of weekly recreational physical activity in a self-administered questionnaire. The questionnaire reports of light, moderate and vigorous physical activity were converted into total kilocalories per week of energy expenditure. A decrease in risk of breast cancer was found with increasing levels of total recreational physical activity (P (trend) = 0.09). The adjusted odds ratio for those women who expended more than 4,000 kcal/week was 0.73 (95% confidence interval (CI) 0.51-1.05) compared with women who undertook no physical activity. The reduction in risk with recreational physical activity was most evident for women who undertook any vigorous activity. These results provide some support for the hypothesis that physical activity may decrease the risk of developing breast cancer.