Background The aim of this study was to investigate stability of physical activity from childhood and adolescence to adulthood in multiple age cohorts, and analyze how well adult physical activity can be predicted by various physical activity variables measured in childhood and adolescence. Methods The data were drawn from the Cardiovascular Risk in Young Finns Study. The study was started in 1980, when cohorts of randomly sampled boys and girls aged 3, 6, 9, 12, 15, and 18 years (total of 2309 subjects) were examined for the first time. The measurements were repeated in 1983, 1986, 1989, 1992, and 2001. In 2001, the subjects (n =1563, 68%) were aged 24, 27, 30, 33, 36, and 39 years, respectively. Physical activity was measured by means of a short self-report questionnaire that was administered individually in connection with a medical examination. On the basis of a questionnaire, a physical activity index (PAI) was calculated. There were no significant differences in the 1980 PAI between participants and dropouts in 2001. Results Spearman℉s rank order correlation coefficients for the 21-year tracking period varied from 0.33 to 0.44 in males, and from 0.14 to 0.26 in females. At shorter time intervals the correlation was higher. On average, the tracking correlation was lower in females than in males. Persistent physical activity, defined as a score in the most active third of the PAI in two or three consecutive measurements, increased the odds that an individual would be active in adulthood. Odds ratios for 3-year continuous activity versus continuous inactivity varied from 4.30 to 7.10 in males and 2.90 to 5.60 in females. The corresponding odds ratios for 6-year persistence were 8.70 to 10.80 and 5.90 to 9.40. Conclusions It was concluded that a high level of physical activity at ages 9 to 18, especially when continuous, significantly predicted a high level of adult physical activity. Although the correlations were low or moderate, we consider it important that school-age physical activity appears to influence adult physical activity, and through it, the public health of the general population. The aim of this study was to investigate stability of physical activity from childhood and adolescence to adulthood in multiple age cohorts, and analyze how well adult physical activity can be predicted by various physical activity variables measured in childhood and adolescence. The data were drawn from the Cardiovascular Risk in Young Finns Study. The study was started in 1980, when cohorts of randomly sampled boys and girls aged 3, 6, 9, 12, 15, and 18 years (total of 2309 subjects) were examined for the first time. The measurements were repeated in 1983, 1986, 1989, 1992, and 2001. In 2001, the subjects (n =1563, 68%) were aged 24, 27, 30, 33, 36, and 39 years, respectively. Physical activity was measured by means of a short self-report questionnaire that was administered individually in connection with a medical examination. On the basis of a questionnaire, a physical activity index (PAI) was calculated. There were no significant differences in the 1980 PAI between participants and dropouts in 2001. Spearman℉s rank order correlation coefficients for the 21-year tracking period varied from 0.33 to 0.44 in males, and from 0.14 to 0.26 in females. At shorter time intervals the correlation was higher. On average, the tracking correlation was lower in females than in males. Persistent physical activity, defined as a score in the most active third of the PAI in two or three consecutive measurements, increased the odds that an individual would be active in adulthood. Odds ratios for 3-year continuous activity versus continuous inactivity varied from 4.30 to 7.10 in males and 2.90 to 5.60 in females. The corresponding odds ratios for 6-year persistence were 8.70 to 10.80 and 5.90 to 9.40. It was concluded that a high level of physical activity at ages 9 to 18, especially when continuous, significantly predicted a high level of adult physical activity. Although the correlations were low or moderate, we consider it important that school-age physical activity appears to influence adult physical activity, and through it, the public health of the general population.