PURPOSE: We identified baseline demographic and clinical factors associated with treatment failure after surgical treatment of stress urinary incontinence. MATERIALS AND METHODS: Data were obtained from 655 women randomized to Burch colposuspension or autologous rectus sling. Of those, 543 (83%) had stress failure status assessed at 24 months (269 Burch, 274 sling). Stress failure (261) was defined as self-report of stress urinary incontinence by the Medical, Epidemiological, and Social Aspects of Aging questionnaire, positive stress test or re-treatment for stress urinary incontinence. Nonstress failure (66) was defined as positive 24-hour pad test (more than 15 ml) or any incontinent episodes by 3-day voiding diary with none of the 3 criteria for stress failure. Subjects not meeting any failure criteria were considered a treatment success (185). Adjusting for surgical treatment group and clinical site, logistic regression models were developed to predict the probability of treatment failure. RESULTS: Severity of urge incontinence symptoms (p = 0.041), prolapse stage (p = 0.013), and being postmenopausal without hormone therapy (p = 0.023) were significant predictors for stress failure. Odds of nonstress failure quadrupled for every 10-point increase in Medical, Epidemiological, and Social Aspects of Aging questionnaire urge score (OR 3.93 CI 1.45, 10.65) and decreased more than 2 times for every 10-point increase in stress score (OR 0.36, CI 0.16, 0.84). The associations of risk factors and failure remained similar regardless of surgical group. CONCLUSIONS: Two years after surgery, risk factors for stress failure are similar after Burch and sling procedures and include greater baseline urge incontinence symptoms, more advanced prolapse, and menopausal not on hormone replacement therapy. Higher urge scores predicted failure by nonstress specific outcomes.