A s 2.5% of births in the United States are impacted by uterine atony, identifying high-risk women can promote timely treatment and subsequently minimize adverse outcomes. However, there is no current risk assessment model that solely focuses on the risk for uterine atony rather than the overall risk for postpartum hemorrhage (PPH). This retrospective study created and tested 2 risk prediction models for antepartum and intrapartum uterine atony.The study pulled data from the Vanderbilt Research Synthetic Derivative, a database with vaginal and cesarean delivery information between 2010 and 2019. A total of 28,706 patients were included (4772 and 23,993 cases of uterine atony and nonuterine atony, respectively). Patients diagnosed with uterine atony during the peripartum period were compared with patients without uterine atony to evaluate risk factors related to demographics, maternal comorbidities, pregnancy-related and labor-related conditions, and delivery-related factors. Parity, gestational age, anemia, macrosomia, epidural analgesia, medication exposures, and race or ethnicity were not included risk factors due to the retrospective nature of the study. The authors evaluated the presence of supplemental uterotonic medication, methylergonovine, and carboprost in cases of uterine atony.Statistical analysis additionally differed between the antepartum model, centered around demographic, maternal morbidity, and pregnancy-related characteristics, and the intrapartum model, based on labor and delivery characteristics. Statistical analyses included standard descriptive statistics and absolute standardized mean difference for demographic and clinical characteristic comparisons across groups, logistic regression with up to 366 degrees of freedom, odds ratios, 95% CIs, χ 2 statistic for group and subgroup comparisons, and the bootstrap method for internal model validation and calibration.Twenty total risk factors were analyzed in both the antepartum and intrapartum models.Multiple gestation, obesity, polyhydramnios, placental abruption, placenta previa, prior cesarean delivery, severe preeclampsia, leiomyomas, prior PPH, maternal age, and diabetes mellitus were antepartum factors that increased the risk of uterine atony in order from greatest to least impact. Intrapartum factors that increased the risk of uterine atony (from greatest to least impact) were chorioamnionitis, multiple gestations, failed induction of labor, polyhydramnios, obesity, placenta previa, cesarean delivery, labor length, placental abruption, prior PPH, maternal age, severe preeclampsia, and leiomyomas. The perineal laceration was not associated with an increased risk of uterine atony.Comparisons between models demonstrated that the intrapartum model was superior at predicting uterine atony compared to the antepartum model. The next steps include application in prospective studies as well as external validation.