Objective To compare the efficacy and safety of endoscopic variceal ligation (EVL) and β-receptor blockers (βB) administration for prophylaxis of primary esophageal variceal bleeding. Methods EMBASE, PubMed, CENTRAL, Wan-fang Database, CNKI database and Google Scholar were comprehensively searched in May 2015 for eligible full-text randomized controlled trials (RCTs). Endpoints of interest were first variceal bleeding, all-cause mortality, bleeding-related mortality and main adverse events. A fixed-effect model was firstly utilized to calculate the pooled odds risk (OR) with 95% confidence intervals (CIs). Results Fourteen studies involving 1 280 patients were identified fulfilled the inclusion criteria in this analysis. Compared with βB, EVL significantly reduced the incidence rate of first variceal bleeding (OR=0.67, 95%CI: 0.49-0.91,P=0.010, I²=21%). All-cause and bleeding-related mortality had no significant difference in the two groups (all-cause mortality [OR=1.14, 95%CI: 0.87-1.50, P=0.35, I²=0%] and bleeding-related mortality [OR=0.72, 95%CI: 0.44-1.17, P=0.19, I²=0%]). EVL had lower incidence rate of main adverse events than non-selective βB (OR=0.43, 95%CI: 0.29-0.64, P<0.000 1, I²=40%). However, compared with EVL, the incidence rate of main adverse events in Carvedilol was significantly lower (OR=10.59, 95%CI: 2.37-47.27, P=0.002). Conclusion EVL is more effective in preventing the primary esophageal variceal bleeding compared with βB, but the two methods have no differences in all-cause mortality or bleeding-related mortality.