The 12-year prediction of physical activity level and submaximal aerobic fitness observed during young adulthood was quantified from childhood and parental measures. The sample consisted of 153 children and adolescents of the Quebec Family Study who were evaluated at baseline and reassessed 12 years later, as young adults. Physical work capacity at a heart rate of 150 beats x min(-1) (PWC150) was measured using cycle ergometry. A 3-day activity record was used to estimate daily energy expenditure (DEE), inactive time (IA), and time spent in moderate-to-vigorous activity (MVPA). Spearman partial inter-age correlations, controlling for length of follow-up and age at baseline, indicated better tracking for PWC150 (0.24 and 0.46, males and females, respectively) than for indicators of physical activity (0.07 < or = r < or = 0.25, males; 0.06 < or = r < or = 0.22, females). Multiple regression analyses indicated that parental measurements of activity and PWC150 did not add any predictive value, with the exception of paternal DEE, which accounted for 8% of the variance in males. There is moderately high stability of submaximal work capacity and lower tracking of physical activity from childhood and adolescence into young adulthood.