Background It has been suggested that EUS-BD may be a feasible and safer alternative to percutaneous transhepatic biliary drainage (PTBD) after failed ERCP in patients with ascites. To date, no study has specifically evaluated the performance of EUS-BD in this context. Methods Retrospective analysis was done for patients with and without ascites who underwent EUS-BD for malignant biliary obstruction after failed ERCP between July 2010 and September 2014. Complications and technical and clinical successes between the two groups were compared. Results A total of 31 patients were included: 20 patients without ascites (group 1) and 11 with ascites (group 2). Nineteen patients underwent EUS-hepaticogastrostomy (six in group 2), and 12 underwent EUS-choledochoduodenostomy (five in group 2). Technical success was achieved in all patients. Clinical success was observed in 95% ( n  = 19) in group 1 and 64% ( n  = 7) in group 2 ( p  = 0.042). In three out of four patients without clinical success in group 2, the follow-up period was not long enough to observe the clinical response because of early death within the 2 weeks after EUS-BD secondary to disease progression or preprocedural unresponsive sepsis. No significant differences were observed between groups 1 and 2 either in the overall rates of procedural-related complications (20 and 9%, respectively, p  = 0.63) or in the rates of major complications (15 vs 9%, respectively, p  = 0.639). Stent migration occurred in one patient in each group, intra- or post-procedural bleeding occurred in two patients in group 1, which was conservatively managed, and one patient in group 1 presented biliary leakage. Stent patency and the number of re-interventions were not significantly different. Conclusions EUS-BD is technically feasible in patients with ascites. Our results suggest that EUS-BD may be a clinically effective and safe alternative after failed ERCP in patients with ascites.