To determine whether women engaged in strenuous, provocative exercise are more likely to be incontinent in future life than similarly fit women who participated in less provocative exercise. In this retrospective cohort study, female American Olympians who competed in swimming (low-impact group) and in gymnastics and track and field (high-impact group) between 1960 and 1976 completed a structured questionnaire. Primary outcome measures included the prevalence of the symptoms of stress and urge incontinence. Statistical analyses of results included χ2, Fisher exact test, two-tailed t tests, Wilcoxon rank sum test, and stepwise multiple logistic regression. P < .05 was considered significant. One hundred four women responded (response rate 51.2%). High-impact athletes were older (46.2 compared with 42.4 years) and were more likely to report incontinence when they were doing their sport as Olympians (35.8% compared with 4.5%) than low-impact athletes; low-impact athletes were more likely to be parous (83.3% compared with 60.7%). There was no difference in the prevalence of the symptom of stress incontinence between the highversus low-impact groups: any incontinence, 41.1% compared with 50%; daily or weekly incontinence, 10.7% compared with 8.3%; and incontinence that bothered them moderately or greatly, 10.7% compared with 4.2%. With our sample size, this study had 80% power to detect a fourfold difference in daily or weekly incontinence between groups, but only a 30% power to detect a twofold difference, given a baseline prevalence of 10%. When age, body mass index (BMI), parity, Olympic sport group, and incontinence during Olympic sport were entered into stepwise logistic regression analyses, only BMI was significantly associated with regular stress or urge incontinence symptoms. Participation in regular, strenuous, highimpact activity when younger did not predispose women to a markedly higher rate of clinically significant urinary incontinence in later life.