Background and study aims  Although endoscopic retrograde cholangiopancreatography (ERCP) is standard of care for malignant biliary obstruction, endoscopic ultrasound-guided biliary drainage (EUS-BD) as a primary treatment has become increasingly utilized. The aim of this study was to perform a systematic review and meta-analysis to evaluate the effectiveness and safety of EUS-BD for primary treatment of malignant biliary obstruction and comparison to traditional ERCP. Methods  Individualized search strategies were developed through November 2018 using PRISMA and MOOSE guidelines. A cumulative meta-analysis was performed by calculating pooled proportions. Subgroup analysis was performed for studies comparing EUS-BD versus ERCP. Heterogeneity was assessed with Cochran Q test or I 2 statistics, and publication bias by funnel plot and Egger's tests. Results  Seven studies (n = 193 patients; 57.5 % males) evaluating primary EUS-BD for malignant biliary obstruction were included. Mean age was 67.4 years (2.3) followed an average of 5.4 months (1.0). For primary EUS-BD, pooled technical success, clinical success, and adverse event (AE) rates were 95 % (95 % CI 91 - 98), 97 % (95 % CI 93 - 100), and 19 % (95 % CI 11 - 29), respectively. Among EUS-BD and ERCP comparator studies, technical and clinical success, and total AEs were not different with lower rates of post-ERCP pancreatitis and reintervention among the EUS-BD group. Conclusion  Primary EUS-BD is an effective treatment with few AE. Comparing EUS-BD versus ERCP, EUS-BD has comparable efficacy and improved safety as a primary treatment for malignant biliary obstruction. Further randomized trials should be performed to identify patient populations and clinical scenarios in which primary EUS-BD would be most appropriate.