BACKGROUND:Questionnaires assessing accidental bowel leakage lack important patient-centered symptoms. OBJECTIVE:We aimed to create a valid measure of accidental bowel leakage symptoms. DESIGN:We previously created a conceptual framework capturing patient-centered accidental bowel leakage symptoms. The framework included bowel leakage type, severity and bother and ancillary bowel symptoms, including predictability, awareness, leakage control, emptying disorders and discomfort. SETTINGS:Outpatient clinics. PATIENTS:Women with at least monthly accidental bowel leakage. INTERVENTIONS:Participants completed the Accidental Bowel Leakage Evaluation at baseline, 12 and 24 weeks, as well as bowel diaries, and other validated pelvic floor questionnaires. A subset completed items twice before treatment. Final item selection was based on psychometric properties and clinical importance. MAIN OUTCOME MEASURES:Psychometric analyses included Cronbach's alphas, confirmatory factor, and item response theory analyses. Construct validity was based on correlations with measures of similar constructs. RESULTS:A total of 296 women completed baseline items and 70 provided test-retest data. The cohort was predominately White (79%) and middle aged (64 +/- 11 years). Confirmatory factor analyses supported the conceptual framework. The final 18-item scale demonstrated good internal consistency (Cronbach's alpha=0.77 -0.90) and test-retest reliability (intraclass correlation=0.80). Construct validity was demonstrated with baseline,12- and 24-week scale scores which correlated with the Vaizey (r=0.52, 0.68 and 0.69), Colo-rectal Anal Distress Inventory (r=0.54, 0.65, 0.71), Colo-rectal Anal Impact Questionnaire (r=0.48, 0.53, 0.53), and hygiene (r=0.39, 0.43, 0.49) and avoidance subscales scores of the adaptive index (r=0.45, 0.44, 0.43) and average number of pad changes per day on bowel diaries (r=0.35, 0.38, 0.31), all p<.001. LIMITATIONS:Validation in a care-seeking population. CONCLUSIONS:The Accidental Bowel Leakage Evaluation instrument is a reliable, patient-centered measure with good validity properties. This instrument improves on currently available measures by adding patient- important domains of predictability, awareness, control, emptying and discomfort. See Video Abstract at http://links.lww.com/DCR/B172.