The traditional management of postoperative rebleeding after upper gastrointestinal haemorrhage is emergency haemostasis by gastroscopy or reoperative exploration [1, 2]. However, such treatment is poorly effective in stopping bleeding, and it is often difficult to identify the bleeding site for reoperation, and surgical hemostasis is extremely difficult due to local adhesions and gastrointestinal wall edema. Patients often cannot tolerate reoperation in the short term, and serious postoperative complications occur. Therefore, exploring new methods that enable timely diagnosis and effective hemostasis is an important topic of clinical research. We used a Toshiba 1250mAC-arm digital subtraction angiography (DSA) machine from Japan. A 5Fcobra or 3Fsp microcatheter was inserted via the right femoral artery using the Seldingers technique. After doing abdominal arteriography, super-selective arteriography was then performed according to different results, and embolization was performed after confirming the site of the bleeding artery. The embolic agents are mainly stainless steel rings with wool and gelatin sponge particles. Selective abdominal arteriography, which revealed the site of bleeding, showed spillage of contrast agent; when the angiogram did not reveal a clear site of bleeding, the indwelling arterial catheter was continuously and slowly injected with hemostatic drugs back to the ward for observation, and when the symptoms of upper gastrointestinal bleeding appeared again immediately the 2nd angiogram revealed the site of bleeding, and both obtained successful hemostasis. The key to the diagnosis and treatment of upper gastrointestinal hemorrhage is how to quickly identify the bleeding site for effective hemostasis. We applied selective abdominal arteriography combined with interventional embolization therapy and achieved good results. I have learned that this therapy is easy to operate, with little damage, and can be tolerated by weak patients, and does not require reopening surgery, so it is easy to get the patients’ approval. Selective abdominal arteriography has accurate qualitative and local diagnostic value for upper gastrointestinal bleeding, and it can show contrast spillage if the bleeding rate is above 05 ml/s. At the same time, it can also show vascular and blood flow abnormalities in the lesion. The application of DSA to show the bleeding artery and determine the safe embolization site is more sensitive than conventional angiography to show spillage and intra-luminal contrast diffusion. Interventional embolization treatment is based on the arteriogram to determine the bleeding site, and targeted injection of drugs and embolic agents directly into the bleeding artery, which can receive good results of immediate hemostasis if done properly. Experience shows that the application of hemostatic drugs has a better effect on smaller vessels and those with extensive bleeding, while vessels with large bleeding are treated with embolization, which has a fast and long-lasting effect. The results show that selective abdominal arteriography is a safe, easy, and reliable screening method for patients with rebleeding after upper gastrointestinal hemorrhage, while interventional embolization is an effective and less invasive way to stop bleeding and can rapidly improve the patient’s bleeding symptoms. There are several aspects of clinical treatment that can affect the efficacy of interventional hemostasis. The ability to accurately visualize the bleeding site and place the appropriate size embolic material is the key to interventional hemostasis. The first step is to capture the timing of the bleeding for imaging. We note that when selectively embolizing the bleeding foci in the superior pancreaticoduodenal artery, it is necessary to superselectively embolize the anterior and posterior superior pancreaticoduodenal arteries respectively, and if necessary, it is necessary to superselectively embolize the anterior and posterior inferior pancreaticoduodenal arteries via the superior mesenteric artery in order to achieve hemostasis. Pay attention to exclude intestinal contents, intestinal gas overlap and motility artifact interference, and pay attention to venous bleeding. Venous bleeding in the upper gastrointestinal tract is difficult to be detected by arteriography, so in addition to the color and speed of bleeding that should be noted from a clinical point of view, positive findings can be obtained by extending the film time to more than 25-30s in the imaging technique, and the DSA acquisition time is approximately the same. It has been reported in the literature that interventional embolization has the risk of post-infarction perforation due to improper operation in the treatment of cavity organ bleeding [4]. However, with the continuous improvement of interventional radiological techniques and catheter materials, successful reports of embolization for upper gastrointestinal bleeding at home and abroad have been increasing in recent years. We believe that: there is a rich vascular network in the wall of the gastroduodenum, which forms collateral anastomoses with each other, and embolization of its arteries or terminal arteries within a certain range is also less likely to result in infarction of the gastrointestinal wall; in the specific operation, the embolic agent should be accurately delivered to the vessels at the bleeding site to avoid entering the splenic artery causing partial splenic necrosis; the embolic agent should be placed in the terminal vessels of the bleeding as far as possible, avoiding placing it in the large vessels that are still far from the bleeding site, because the presence of collateral The presence of anastomosis can affect the hemostatic effect; gelatin sponge particles can be absorbed in a short period of time and rebleeding occurs, so avoid using gelatin sponge particles alone and add stainless steel rings with wool. Nevertheless, the application of interventional embolization therapy has its limitations, such as extensive gastric mucosal bleeding and venous bleeding from the portal venous system, where the hemostatic effect is poor. Therefore, we are required to correctly apply the interventional embolization operation technique to treat upper gastrointestinal hemorrhage from the specific clinical situation.