Reliable data are scanty on the incidence of chronic diseases and life expectancy (LE) of highly trained athletes. We therefore studied Finnish male world class athletes to estimate the LE of athletes. Finnish team members in the Olympic games, World or European championships or intercountry competitions during 1920-1965 in track and field athletics, cross-country skiing, soccer, ice hockey, basketball, boxing, wrestling, weight lifting, and shooting were included (N = 2613 men). The reference cohort, 1712 men, was selected from the Finnish Defence Forces conscription register matched on age and area of residence. All referents were classified completely healthy at the time of induction to military service. The stratified Kaplan-Meier product limit method and the Cox proportional hazards model were used to estimate the life expectancies and the mortality odds ratios (OR) and their confidence limits. The mean LE adjusted for occupational group, marital status, and the age at entry to the cohort (and its 95% confidence limits) was in endurance sports (long distance running and cross-country skiing) 75.6 (73.6, 77.5) yr; in team games (soccer, ice hockey, basketball, as well as jumpers and short-distance runners from track and field (73.9 (72.7, 75.1) yr; in power sports (boxing, wrestling, weight lifting, and throwers from field athletics) 71.5 (70.4, 72.2) yr; and in the reference group 69.9 (69.0, 70.9) yr. The increased mean life expectancies were mainly explained by decreased cardiovascular mortality (endurance sports mortality odds ratio OR = 0.49 (95% CL 0.26, 0.93), team sports OR = 0.61 (0.41, 0.92) compared with referents). For maximum life span no differences between the groups were observed.(ABSTRACT TRUNCATED AT 250 WORDS)