The aim of this study is to explore the clinical application value of preoperative selective partial splenic embolization before splenectomy plus portal-azygous disconnection. 158 cases of liver cirrhosis combined with upper gastrointestinal hemorrhage patients were selected, which were randomly divided into splenic embolization group (S, n=77) and the non-splenic embolization group (U, n=81). Group S patients were firstly performed partial splenic embolization (PSE), and then underwent splenectomy plus portal-azygous disconnection, and the group U patients were directly performed splenectomy plus portal-azygous disconnection. Statistical analysis was used SAS8.0 statistical analysis software. One week after partial splenic embolization, the platelet of group S returned to normal, and the rise of white blood cells and hemoglobin, and shorten of prothrombin time in group S were much better than that in group U (P < 0.01, P < 0.05); the indexes, such as the intraoperative blood loss, the blood transfusion amount, the amount of platelet infusion, and the incidence of complications in group S were more superior than that in group U (P < 0.05). Preoperative selective splenic artery embolization before splenectomy plus portal-azygous disconnection can restore the spleen function, and reduce the risk of surgery and incidence of complications.