OBJECTIVE: To investigate the association between parity and urinary incontinence, including subtypes and severity of incontinence, in an unselected sample, with special emphasis on age as a confounder or effect modifier. METHODS: This was a cross-sectional study (response rate 80%) with 27,900 participating women. Data on parity and urinary leakage, type, frequency, amount, and impact of incontinence were recorded by means of a questionnaire. A validated severity index was used. Relative risks (RR) with nulliparous women as reference were used as an effect measure. RESULTS: Incontinence was reported by 25% of participants. Prevalences among nulliparous women ranged from 8% to 32%, increasing with age. Parity was associated with incontinence, and the first delivery was the most significant. The association was strongest in the age group 20–34 years with RR 2.2 (95% confidence interval [CI] 1.8, 2.6) for primiparous women and 3.3 (2.4, 4.4) for grand multiparous women. A weaker association was found in the age group 35–64 years (RRs between 1.4 and 2.0), whereas no association was found among women over 65 years. For stress incontinence in the age group 20–34 years, the RR was 2.7 (2.0, 3.5) for primiparous women and 4.0 (2.5, 6.4) for grand multiparous women. There was an association with parity also for mixed incontinence, but not for urge incontinence. Severity was not clinically significantly associated with parity. CONCLUSION: Parity is an important risk factor for female urinary incontinence in fertile and peri- and early postmenopausal ages. Only stress and mixed types of incontinence are associated with parity. All effects of parity seem to disappear in older age.