To estimate obstetric risk factors of fecal incontinence among middle-aged women. We conducted a mail survey of the Gazel cohort of volunteers for epidemiologic research. In 2000, a questionnaire on anal incontinence was mailed to 3,114 women who were then between the ages of 50 and 61 years; 2,640 (85%) women returned the completed questionnaire. Fecal incontinence was defined by involuntary loss of stool. Logistic regression was used to estimate the effect of obstetric and general risk factors. Prevalence of fecal incontinence in the past 12 months was 9.5% (250). Significant risk factors for fecal incontinence were completion of high school (adjusted odds ratio [OR] 1.5, 95% confidence interval [CI] 1.1-2.0), self-reported depression (OR 2.1, 95% CI 1.6-2.7), overweight or obesity measured by body mass index (BMI) (OR 1.5 for BMI of 25-30, 95% CI 1.1-2.0; OR 1.6 for BMI more than 30, 95% CI 1.1-2.5), surgery for urinary incontinence (OR 3.5, 95% CI 2.0-6.1), and anal surgery (OR 1.7, 95% CI 1.1-2.9). No obstetric variable (parity, mode of delivery, birth weight, episiotomy, or third-degree perineal tear) was significant. Prevalence of fecal incontinence was similar for nulliparous, primiparous, secundiparous, and multiparous women (11.3%, 9.0%, 9.0%, and 10.4%, respectively), and among parous women, it was similar for women with spontaneous vaginal, instrumental (at least one), or only cesarean deliveries (9.3%, 10.0%, and 6.6%, respectively). In our population of women in their 50s, fecal incontinence was not associated with either parity or mode of delivery.