PURPOSE: To investigate the relationship between baseline leisure-time physical activity and changes in leisure activity during follow-up on long-term weight changes. METHODS: We evaluated prospectively 11,974 participants (university graduates) who participated in a dynamic cohort (Seguimiento Universidad de Navarra cohort) with an average follow-up of 27 months. Self-reported data from validated mailed questionnaires were used. Baseline leisure activity was assessed with a previously validated questionnaire. RESULTS: After adjusting for age, hours sitting down, smoking status, snacking, fiber intake, and consumption of sugar-sweetened beverages, fast food, and alcohol, participants who decreased their leisure activity during follow-up experienced a significant increase in body mass index (BMI; relative change): for men, 0.9% (95% confidence interval [CI] = 0.5-1.2%); for women, 1.0% (95% CI = 0.6-1.3%). Participants who increased their leisure activity during follow-up experienced a significant reduction (relative change) in BMI: for men, -0.8% (95% CI = -1.1% to -0.5%); for women, -0.6% (95% CI = -0.9% to -0.4%). This inverse association between changes in leisure activity and weight gain was significantly stronger for participants with a baseline BMI >or=25 kg x m(-2), but the absolute magnitude of this interaction effect was trivial. Baseline physical activity was not significantly associated with weight changes after 2-yr of follow-up. CONCLUSION: Longitudinal changes in leisure activity during follow-up were inversely associated with changes in body weight. The true relationships between leisure activity and body weight are likely to have been larger than observed, owing to attenuation of effects by measurement error in self-reported data.