A 25-year population-based study was conducted to determine how physical fitness and participation in leisure-time physical activity in adolescence (age 12-18 years) predict leisure-time physical activity in adulthood (age 37-43 years). In 1976, five field tests were conducted to measure muscular fitness, agility and aerobic capacity, and self-report weekly frequencies of activity were obtained by questionnaire. A modified questionnaire was repeated in 2001, when participants were age 37-43 years (N = 1525). On the basis of the questionnaire, a physical activity index was calculated both in 1976 and 2001. The odds ratios (OR) for being inactive in adulthood among those who were physically very active in adolescence compared with those who were inactive in adolescence was 0.13 for males (95%CI: 0.06-0.31) and 0.28 for females (95%CI: 0.13-0.59). The odds ratio for adult inactivity among males in the highest versus lowest fitness group in adolescence was 0.24 (95%CI: 0.07-0.81). Activity in adolescence predicted activity in adulthood in both males and females. The risk for adult inactivity was significantly lower for those who were physically active in adolescence. In addition, high fitness predicted adult activity among males but not among females.