Background/Aims: Gastric varices (GV) are present in 25% of cirrhotic patients with high rates of rebleeding and mortality. Data on endoscopic ultrasound (EUS)-guided treatment in severe liver disease (model for end stage liver disease sodium [MELD-Na] >18 and Child-Turcotte-Pugh [CTP] C with GV) are scarce. Thus, we performed a retrospective comparison of endoscopic glue injection with EUS-guided therapy in cirrhotic patients with large GV. Methods: A retrospective study was performed in the tertiary hospitals of India. A total of 80 patients were recruited. The inclusion criteria were gastroesophageal varices type 2, isolated gastric varices type 1, bleeding within 6 weeks, size of GV >10 mm, and a MELD-Na >18. Treatment outcomes and complications of endoscopic glue injection and EUS-guided GV therapy were compared. Results: In this study, the patients’ age, sex, liver disease severity (CTP, MELD-Na) and clinical parameters were comparable. The median number of procedures, injected glue volume, complications, and GV obturation were better in the EUS group, respectively. On subgroup analysis of the EUS method (e.g., direct gastric fundus vs. paragastric collateral [PGC] coil placement), PGC coil placement showed decreased coil requirement, less injected glue volume, decreased luminal coil extrusion, and increased successful GV obturation. Conclusions EUS-guided treatment is more efficient and safer, and requires a smaller number of treatment sessions, as compared to endoscopic treatment in severe liver disease patients with large GV. Furthermore, PGC coil placement increases the complete obliteration of GV.