A decline in mortality from cardiovascular disease over the past 30 years has been well documented, but the reasons for the decline remain unclear. We analyzed the 10-year incidence of cardiovascular disease and death from cardiovascular disease in three groups of men who were 50 to 59 years old at base line in 1950, 1960, and 1970 (the 1950, 1960, and 1970 cohorts) in order to determine the contribution of secular trends in the incidence of cardiovascular disease, risk factors, and medical care to the decline in mortality. The 10-year cumulative mortality from cardiovascular disease in the 1970 cohort was 43 percent less than that in the 1950 cohort and 37 percent less than that in the 1960 cohort (P = 0.04 by log-rank test). Among the men who were free of cardiovascular disease at base line, the 10-year cumulative incidence of cardiovascular disease declined approximately 19 percent, from 190 per 1000 in the 1950 cohort to 154 per 1000 in the 1970 cohort (0.10 less than P less than 0.20 by chi-square test), whereas the 10-year rate of death from cardiovascular disease declined 60 percent (relative risk for the 1950 cohort as compared with the 1970 cohort, 2.53; 95 percent confidence interval, 1.22 to 5.97). Significant improvements were found in risk factors for cardiovascular disease among the men initially free of cardiovascular disease in the 1970 cohort as compared with those in the 1950 cohort, including a lower serum cholesterol level (mean +/- SD, 5.72 +/- 0.98 mmol per liter [221 +/- 38 mg per deciliter], as compared with 5.90 +/- 1.03 mmol per liter [228 +/- 40 mg per deciliter]) and a lower systolic blood pressure (mean +/- SD, 135 +/- 19 mm Hg, as compared with 139 +/- 25 mm Hg), better management of hypertension (22 percent vs. 0 percent were receiving antihypertensive medication), and reduced cigarette smoking (34 percent vs. 56 percent). We propose that these improvements may have had more pronounced effects on mortality from cardiovascular disease than on the incidence of cardiovascular disease in this population. Our data suggest that the improvement in cardiovascular risk factors in the 1970 cohort may have been an important contributor to the 60 percent decline in mortality in that group as compared with the 1950 cohort, although a decline in the incidence of cardiovascular disease and improved medical interventions may also have contributed to the decline in mortality.