Background: The protective effect of physical activity (PA) on abdominal adiposity is unclear. Objective: We examined whether PA independently predicted gains in body weight and abdominal adiposity. Design: In a prospective cohort study [the EPIC (European Prospective Investigation into Cancer and Nutrition)], we followed 84,511 men and 203,987 women for 5.1 y. PA was assessed by a validated questionnaire, and individuals were categorized into 4 groups (inactive, moderately inactive, moderately active, and active). Body weight and waist circumference were measured at baseline and self-reported at follow-up. We used multilevel mixed-effects linear regression models and stratified our analyses by sex with adjustments for age, smoking status, alcohol consumption, educational level, total energy intake, duration of follow-up, baseline body weight, change in body weight, and waist circumference (when applicable). Results: PA significantly predicted a lower waist circumference (in cm) in men (beta = -0.045; 95% CI: -0.057, -0.034) and in women (beta = -0.035; 95% CI: -0.056, -0.015) independent of baseline body weight, baseline waist circumference, and other confounding factors. The magnitude of associations was materially unchanged after adjustment for change in body weight. PA was not significantly associated with annual weight gain (in kg) in men (beta = -0.008; 95% CI: -0.02, 0.003) and women (beta = -0.01; 95% CI: -0.02, 0.0006). The odds of becoming obese were reduced by 7% (P <0.001) and 10% (P <0.001) for a one-category difference in baseline PA in men and women, respectively. Conclusion: Our results suggest that a higher level of PA reduces abdominal adiposity independent of baseline and changes in body weight and is thus a useful strategy for preventing chronic diseases and premature deaths.