BACKGROUND: As epidemiological studies have become more complex, demands for short, easily administered measures of risk factors have increased. This study investigates whether such a measure of physical activity is associated with the risk of death from all causes and death from specific causes. METHODS: A prospective follow-up study of 11,090 men and women, aged 35-64 years, recruited from five UK general practices who responded to a postal questionnaire in 1989. Self-reported frequency of vigorous-intensity physical activity and data on confounding factors were collected at baseline survey. Death notifications up to 31 December 2001 were provided by the Office for National Statistics. The relative risk (and 95% confidence interval) of dying associated with each level of exposure to physical activity was estimated by the hazard ratio in a series of Cox regression models. RESULTS: After >10 years' follow-up there were 825 deaths among the 10 522 subjects with no previous history of angina or myocardial infarction. Participation in vigorous exercise was associated with a significantly lower risk of all-cause mortality. Similar associations were found for ischaemic heart disease and cancer mortality, although the relationships were not significant at the 5% level. CONCLUSIONS: Simple measures of self-reported vigorous physical activity are associated with the risk of future mortality, at least all-cause mortality in a somewhat selected group. Interpretation of the finding should be treated with caution due to the reliance on self-report and the possibility that residual confounding may underlie the associations. Because moderate-intensity physical activity is also beneficial to health, short physical activity questionnaires should include measures of such physical activity in the future.