Cardiovascular events over 24 years of surveillance were examined in 1166 men participating in the Framingham Study, classified by physical demands of their work and by a 24-hour index of physical activity. Findings are based on 303 noncardiovascular, 220 coronary, and 325 cardiovascular deaths in men aged 45 to 64 years at time of physical activity assessment. For level of physical activity over 24 hours, there is a clear trend of improved overall, cardiovascular, and coronary mortality with increased level of physical activity at all ages, including the elderly. The effect is sustained with a more pronounced effect with the passage of time, despite presumed decrease in level of activity. The mortality benefits apply both in those with and without intervening overt cardiovascular disease, making it unlikely that the physical inactivity-mortality relationships reflects already existent myocardial damage. For physical demands of the job, there is only a suggestion (not statistically significant) of benefit for cardiovascular mortality including coronary deaths. In sharp contrast, noncardiovascular mortality was positively related to both physical demand of the job and 24-hour physical activity index.