We studied physical fitness and physical activity in relation to all-cause and cancer mortality in a cohort of 7080 women and 25,341 men examined at the Cooper Clinic in Dallas, Texas, during 1970 to 1989. Physical fitness was assessed at baseline by a maximal treadmill exercise test, while physical activity was self-reported on the attendant health habits questionnaire. Both men and women averaged about 43 years of age at baseline (range, 20 to 88 years), and they were followed for approximately 8 years on average. Through the end of 1989, the women contributed 52,982 person-years of observation and incurred 89 deaths, including 44 deaths due to cancer. The men contributed 211,996 person-years and incurred 601 deaths, with 179 due to cancer. After adjustment for baseline differences in age, examination year, cigarette habit, chronic illnesses, and electrocardiogram abnormalities, we found a strong inverse association between risk of all-cause mortality and level of physical fitness in both men and women (P for trend < 0.001). Physically active men also were at lower risk of all-cause mortality than were sedentary ones (P for trend = 0.01). Among women, however, self-reported physical activity was not significantly related to risk of death from all causes. The risk of mortality from cancer declined sharply across increasing levels of fitness among men (P for trend < 0.001), whereas among women the gradient was suggestive but not significant (P for trend = 0.07). Physically active men also were at lower risk of death from cancer than were sedentary men (P for trend = 0.002), but among women physical activity was unrelated to cancer mortality.