Reported prevalence estimates for fecal incontinence among community-dwelling adults vary widely. A systematic review was undertaken to investigate the studied prevalence of fecal incontinence in the community and explore the heterogeneity of study designs and sources of bias that may explain variability in estimates. A predetermined search strategy was used to locate all studies published that reported the prevalence of fecal incontinence in a community-based sample of adults. Data were extracted onto a proforma for sampling frame and method, sample size, response rate, definition of fecal incontinence used, data-collection method, and prevalence rates. Included studies were critically appraised for possible sources of selection bias, information bias, and imprecision. A total of 16 studies met the inclusion criteria. These could be grouped into definitions of incontinence that included or excluded incontinence of flatus. The estimated prevalence of anal incontinence (including flatus incontinence) varied from 2 to 24 percent, and the estimated prevalence of fecal incontinence (excluding flatus incontinence) varied from 0.4 to 18 percent. Only three studies were found to have a study design that minimized significant sources of bias, and only one of these used a validated instrument for data collection. The prevalence estimate of fecal incontinence from these studies was 11 to 15 percent. No pooling of estimates was undertaken because there was wide variation in study design. A consensus definition of fecal incontinence is needed that accounts for alterations in quality of life. Further cross-sectional studies are required that minimize bias in their design and use validated self-administered questionnaires.