Objective To estimate differences in pelvic floor disorders by mode of delivery. Methods We recruited 1,011 women for a longitudinal cohort study, 5-10 years after first delivery. Using hospital records, we classified each birth as: cesarean without labor, cesarean during active labor, cesarean after complete cervical dilation, spontaneous vaginal birth, or operative vaginal birth. At enrollment, stress incontinence, overactive bladder, anal incontinence, and prolapse symptoms were assessed with a validated questionnaire. Pelvic organ support was assessed using the Pelvic Organ Prolapse Quantification system. Logistic regression analysis was used to estimate the relative odds of each pelvic floor disorder by obstetric history, adjusting for relevant confounders. Results Compared to cesarean without labor, spontaneous vaginal birth was associated with a significantly greater odds of stress incontinence (odds ratio (OR) 2.9; 95% confidence interval (CI) 1.5, 5.5) and prolapse to or beyond the hymen (OR 5.6; 95% CI 2.2, 14.7). Operative vaginal birth significantly increased the odds for all pelvic floor disorders, especially prolapse (OR 7.5; 95% CI 2.7, 20.9). These results suggest that 6.8 additional operative births or 8.9 spontaneous vaginal births, relative to cesarean births, would lead to one additional case of prolapse. Among women delivering exclusively by cesarean, neither active labor nor complete cervical dilation increased the odds for any pelvic floor disorder considered, although the study had less than 80% power to detect a doubling of the odds with these exposures. Conclusion Although spontaneous vaginal delivery was significantly associated with stress incontinence and prolapse, the most dramatic risk was associated with operative vaginal birth.