Introduction Large practice variation exists in mode of delivery after cesarean section, suggesting variation in implementation of contemporary guidelines. We aim to evaluate this practice variation and to what extent this can be explained by risk factors at patient level. Material and methods This retrospective cohort study was performed among 17 Dutch hospitals in 2010. Women with one prior cesarean section without a contraindication for a trial of labor were included. We used multivariate logistic regression analysis to develop models for risk factor adjustments. One model was derived to adjust the elective repeat cesarean section rates; a second model to adjust vaginal birth after cesarean rates. Standardized rates of elective repeat cesarean section and vaginal birth after cesarean per hospital were compared. Pseudo‐R2 measures were calculated to estimate the percentage of practice variation explained by the models. Secondary outcomes were differences in practice variation between hospital types and the correlation between standardized elective repeat cesarean section and vaginal birth after cesarean rates. Results In all, 1068 women had a history of cesarean section, of whom 71% were eligible for inclusion. A total of 515 women (67%) had a trial of labor, of whom 72% delivered vaginally. The elective repeat cesarean section rate at hospital level ranged from 6 to 54% (mean 29.8, standard deviation 11.8%). Vaginal birth after cesarean rates ranged from 50 to 90% (mean 71.8%, standard deviation 11.1%). More than 85% of this practice variation could not be explained by risk factors at patient level. Conclusion A large practice variation exists in elective repeat cesarean section and vaginal birth after cesarean rates that can only partially be explained by risk factors at patient level.