Background: Prognostic scales should be used for early stratification of patients according to risk. Several risk scores have been developed, most of which include endoscopic findings. Objective: To compare three scores "Rockall, Blatchford and AIMS65 scores" to identify the most accurate score used in predicting unfavorable outcomes during patient hospitalization, and for about 1 week after discharge. Patients and methods: A prospective study was conducted on Egyptian patients presented by upper gastrointestinal bleeding, (UGIB) recruited from the emergency department (ED) of Nasser institute hospital for research and treatment, over a period from January 2019 to July 2019. Results: About half of cases had chronic liver disease. The most frequent clinical condition was Melena; present in about half of cases. Stigmata were in more than half of cases. Cases with need for interventional endoscopy had significant higher hepatic disease, severe comorbidities, melena, syncope, shock and worser laboratories& endoscopic findings. Conclusion: Risk stratification and decision to perform interventions including therapeutic endoscopy is often a subjective matter.