Objectives To evaluate the severe outcomes in pregnancies with pernicious placenta previa and prior cesarean delivery, and to analyze the predictive examinations and prevention strategies regarding severe complications in these women. Design Multicentre retrospective cohort study. Setting Thirteen hospitals in China. Population Selected 747 women with pernicious placenta previa and prior cesarean history. Methods The basic data relating to pregnancy, the examinations during gestation and the outcomes were collected and analyzed. Main outcome measures Outcomes including massive bleeding, placenta implantation and hysterectomy were monitored. Results Our results showed that placental implantation occurred in 47.5% of patients and the hysterectomy rate was 10.4%. The incidence of massive bleeding and blood transfusion was 55.8% and 64%, respectively. The women with placenta implantation predicted by both ultrasound and MRI had higher blood loss than those diagnosed only by either one of the examinations. Surprisingly, vascular occlusion had no effect on the occurrence of severe hemorrhage and hysterectomy. The latter was even higher in the vascular blocking group compared to that in the non-blocking group (34.2% and 8.6%, respectively). Conclusions: The pregnancies with pernicious placenta previa and prior cesarean delivery, had a dramatically higher risk of placenta implantation, hysterectomy and massive hemorrhage. MRI combined with ultrasound examination presented high accuracy in predicting severe outcomes in these patients. Vascular occlusion does not appear to be an effective approach to prevent severe outcomes. Surgical hemostasis should be the key goal in blocking massive bleeding, preserving the uterus and improving the prognosis of the patients.