Patients with gastrointestinal bleeding are in a vicious condition, and the resuscitation process requires constant adjustment of the plan according to the changes in their condition. Here we present a successful case and talk about some details. The patient, 50 years old, male, was resuscitated locally for 6 days for gastrointestinal hemorrhage without significant improvement and was referred to our department. According to the medical history, the patient had a history of hepatitis and was considered to be bleeding from post-hepatitis B cirrhosis portal hypertension. The patient was already critically ill on admission, with hematocrit down to less than 5 grams, extremely anemic appearance, massive ascites, poor liver function, and about 20 grams of albumin. On admission, he vomited 1000 ml of fresh blood and had 3 black stools of 300-400 ml of dark red blood each. The patient was on the verge of death due to massive blood loss, gradual decrease in blood pressure, and heart rate of 130 beats/min. What kind of hemostasis method should be used for such a patient with a large amount of bleeding? After treatment, the bleeding was initially stopped. The patient had poor liver function, massive ascites, and also blood loss, what kind of transfusion is appropriate? Blood transfusion is necessary to correct severe anemia, but to what extent is it appropriate? Is a plasma transfusion necessary? The patient was given growth inhibitor, but growth inhibitor is very expensive, is growth inhibitor really an effective drug? One day after the bleeding stopped, the patient bled again, what was the cause of the bleeding? After judgment, I decided to adjust the hemostatic measures to stop the bleeding again, stop the gastrointestinal decompression after a week of successful hemostasis, perform surgical cutoff after 10 days, start eating after 5 days of surgery, and recover and discharge from the hospital 2 weeks after surgery.