A delayed heart rate (HR) recovery after graded exercise testing has been associated with increased all-cause mortality in clinic-based samples. No prior study has examined the association of HR recovery after exercise with the incidence of coronary heart disease (CHD) and cardiovascular disease (CVD) events. We evaluated 2,967 Framingham study subjects (1,400 men, mean age 43 years) who were free of CVD and underwent a treadmill exercise test (Bruce protocol) at a routine examination. We examined the relations of HR recovery indexes (decrease in HR from peak exercise) to the incidence of a first CHD or CVD event and all-cause mortality, adjusting for established CVD risk factors. During follow-up (mean 15 years), 214 subjects experienced a CHD event (156 men), 312 developed a CVD event (207 men), and 167 died (105 men). In multivariable models, continuous HR recovery indexes were not associated with the incidence of CHD or CVD events, or with all-cause mortality. However, in models evaluating quintile-based cut points, the top quintile of HR recovery (greatest decline in HR) at 1-minute after exercise was associated with a lower risk of CHD (hazards ratio vs bottom 4 quintiles 0.54, 95% confidence interval [CI], 0.32 to 0.93) and CVD (hazards ratio 0.61, 95% CI 0.41 to 0.93), but not all-cause mortality (hazards ratio 0.99, 95% CI 0.60 to 1.62). In our community-based sample, HR recovery indexes were not associated with all-cause mortality. A very rapid HR recovery immediately after exercise was associated with lower risk of CHD and CVD events. These findings should be confirmed in other settings.