Objective: To investigate the effects of primary preventive treatment under endoscope for esophageal and gastric varices on bleeding rate and its relevant factors. Methods: 127 cases with liver cirrhosis accompanied with esophageal and gastric varices without bleeding history were included in the endoscopic and non-endoscopic treatment group, respectively. Informed consent was obtained from both groups. Gastric varices (Lgf) and esophageal varices (Leg) were diagnosed according to LDRf classification criteria, and the corresponding treatment scheme was selected according to the recommended principle of this method.The incidence rate of bleeding from ruptured esophageal varices were observed at 3, 6 months, and 1, and 2 years in the treated and the untreated group, and the patients with different Child-Pugh scores were followed-up for 2 years. Gender, age, etiology, varicose degree, Child-Pugh grade, platelet count, prothrombin activity, portal vein thrombosis, collateral circulation, portal vein width and other factors affecting the bleeding rate were assessed. Measurement data were described as mean ± standard deviation (x¯±s), and qualitative data of categorical variables were expressed as percentage (%), and χ2 test was used. Results: 127 cases were followed up for 2 years. There were 55 cases in the endoscopic treatment group (18 cases underwent band ligation, 2 cases underwent band ligation combined with tissue adhesive embolization, 28 cases underwent sclerotherapy, and 7 cases underwent sclerotherapy combined with tissue adhesive embolization). Recurrent bleeding and hemorrhage was occurred in 5 (9.1%) and 28 cases (38.9%), respectively (P<0.05). In addition, there were 72 cases in the untreated group (P<0.05). Severe varicose veins proportions in treated and untreated group were 91.1% and 85.1%, respectively (P>0.05). There was no statistically significant difference in liver cirrhosis-related medication and β-blocker therapy between the treated and untreated group (P>0.05). There was no statistically significant difference in the bleeding rate between the different treated groups (P>0.05). The bleeding rates at 3, 6 months, 1, and 2 years in endoscopic treated and untreated group were 2.00% vs. 2.59% (P>0.05), 2.30% vs. 5.88% (P>0.05), 3.10% vs. 7.55% (P>0.05) and 4.00% vs. 21.62% (P<0.05), respectively. All patients with Child-Pugh grade A, B and C in the treated and the untreated group were followed-up for 2 years, and the bleeding rates were 1.8% vs. 8.1% (P<0.05), 1.1% vs. 9.4% (P<0.05) and 9.1% vs. 10.1% (P>0.05), respectively. There were statistically significant differences in the rupture and bleeding of esophageal and gastric varices, varices degree, Child-Pugh grade and presence or absence of thrombosis formation in portal vein (P<0.05); however, no statistically significant differences in gender, age, etiology, platelet count, prothrombin activity, collateral circulation and portal vein width (P>0.05). There was no intraoperative bleeding and postoperative related serious complications in the treated group. Conclusion: The risk of initial episodes of bleeding from esophageal and gastric varices is significantly correlated with the varices degree, Child-Pugh grade, and portal vein thrombosis. Primary preventive treatment under endoscope is safe and effective for reducing the long-term variceal bleeding risk from esophageal and gastric varices.