Upper gastrointestinal bleeding(UGIB)is one of the most common clinical emergencies.In recent years,due to the aging of the population,the development of gastroscopy technology and the increasing usage of non-steroidal anti-inflammatory drugs and anticoagulant drugs,the etiological ratio of UGIB has gradually changed.In addition,risk assessment for upper gastrointestinal bleeding is constantly being im-proved.At present,Glasgow-Blatchford score(GBS),Rockall score system(RS),AIMS 65 score and PNED score are widely used in clinical practice and have a high degree of recognition,but their best application is slightly different,such as GBS has a high ability to predict the demand for hospital intervention,and classification management through GBS has been shown to reduce hospitalization rate,length of stay and cost utiliza-tion.While RS,AIMS65 score,and PNED score are more suitable for predicting mortality risk,especially in hospitalization mortality,AIMS65 scores seem to perform better.And PNED score is more suitable for 30-day death risk prediction.This article reviews the changes in etiological composition and risk assessment of UGIB in recent years through literature review to understand the relevant research progress.