We studied the adoption or maintenance of physical activity and other optional lifeway patterns for their influence on mortality rates of Harvard College alumni. Men aged 45-84 in 1977, surveyed by questionnaire in 1962 or 1966 and again in 1977, were followed from 1977 through 1988 or to age 90. Of 14,786 alumni, 2,343 died in 165,402 man-years of follow-up. Relative risks of death, standardized for potential confounding influences, for men who between questionnaires increased their physical activity through walking, stair climbing, and sports or recreational activities to 1,500 kcal or more per wk were 0.72 (95% confidence interval 0.64-0.82), compared with 1.00 for men who remained less active. Corresponding relative risks for men who adopted moderately vigorous sports play (> or = 4.5 METs) were 0.73 (0.65-0.81) vs 1.00 for men not adopting such sports; and for cigarette smokers who quit, 0.74 (0.65-0.84) vs 1.00 for persistent smokers. Men with recently diagnosed hypertension had a lower death risk than long-term hypertensives (0.80; 0.70-0.92), as did men with consistent normotension (0.52; 0.47-0.58). Changes in body-mass index had little influence on mortality during follow-up. These findings fit the hypothesis that adopting a physically active lifeway, quitting cigarette smoking, and remaining normotensive independently delay all-cause mortality and extend longevity.