Objective: To establish a prediction model to help the doctor determine which patients are more suitable for transvaginal repair based on the prediction model. Design:All enrolled patients underwent CSD repair performed by a single team. All women in this study had a follow-up clinic visit at 6 months to record their menstruation and measure multiple parameters of the CSD by MRI. Setting:Retrospective study Sample: This study included 1015 women who underwent transvaginal repair of cesarean scar diverticulum (CSD) at Xinhua Hospital and Shanghai First Maternity & Infant Hospital between June 2014 and May 2021. Main outcome measures: CSD patients are categorized as having optimal healing when the menstruation duration is no more than 7 days and the thickness of residual myometrium（TRM） is no less than 5.39 mm after vaginal repair. The final nomogram is constructed to predict surgical outcomes based on pre- and postoperative variables. Results: The key factors determining optimal healing are the timing of cesarean section; menstrual cycle; CSD length, width, depth, and the myometrial layer thickness of the lower uterine segment. With the prediction model, scores are given to each parameter according to the statistics. Total scores range from 0 to 25 points with a cutoff point of 16.5. Predicted that transvaginal repair achieves optimal healing when a score greater than 16.5 points. Uterine position and preoperative TRM are the key factors affecting postoperative TRM. The width of the CSD and the thickness of the lower uterine segment are the key factors affecting postmenstrual abnormal uterine bleeding (P<0.01). Conclusions: We establish a prediction model system for the first time that may predict the repair effect of CSD and can potentially be useful in future clinical trials to determine which patients should be repaired or other treatment options.