Aims The aim of this study was to investigate the impact of persistence and non-persistence in leisure time physical activity on coronary heart disease and all-cause mortality. Methods and results In the Copenhagen City Heart Study, we prospectively followed 12,314 healthy subjects for 33 years of maximum follow-up with at least two repeated measures of physical activity. The association between persistence and non-persistence in leisure time physical activity, coronary heart disease and all-cause mortality were assessed by multivariable Cox regression analyses. Coronary heart disease mortality for persistent physical activity in leisure compared to persistent sedentary activity were: light hazard ratio (HR) 0.76; 95% confidence interval (CI) 0.63-0.92, moderate HR 0.52; 95% CI 0.41-0.67, and high physical activity HR 0.51; 95% CI, 0.30-0.88. The differences in longevity were 2.8 years for light, 4.5 years for moderate and 5.5 years for high physical activity. A substantial increase in physical activity was associated with lower coronary heart disease mortality (HR 0.75; 95% CI 0.52-1.08) corresponding to 2.4 years longer life, whereas a substantial decrease in physical activity was associated with higher coronary heart disease mortality (HR 1.61; 95% CI 1.11-2.33) corresponding to 4.2 years shorter life than the unchanged group. A similar pattern was observed for all-cause mortality. Conclusion We found inverse dose-response relationships between persistent leisure time physical activity and both coronary heart disease and all-cause mortality. A substantial increase in physical activity was associated with a significant gain in longevity, whereas a decrease in physical activity was associated with even greater loss of longevity.