BACKGROUND Cyanoacrylate may form a barrier that prevents bacterial invasion when used in tissue. Because cyanoacrylate polymerizes within seconds on contact with aqueous media, it is used worldwide to arrest gastric variceal bleeding. The aim of this study was to determine the frequency of bacteremia after endoscopic cyanoacrylate injection for gastric variceal bleeding. METHODS Patients with cirrhosis who underwent endoscopic cyanoacrylate injection for gastric variceal bleeding were included. Patients with cirrhosis who underwent upper endoscopy for nonvariceal upper GI bleeding were recruited as controls. Patients with infection before endoscopy were excluded. Blood was cultured in both groups. Injection needles and endoscope accessory channels were cultured in the cyanoacrylate injection group. RESULTS More patients injected with cyanoacrylate had positive blood cultures in comparison with the control group (15/47 vs. 1/47, p < 0.0001). In the cyanoacrylate injection group, the volume of blood transfused and Child-Pugh score were factors associated with the occurrence of bacteremia. Most episodes of bacteremia were transient, except for 1 patient who died of sepsis. Most of the microorganisms cultured from blood samples were identical to those cultured from injection needles (65%) and accessory channels (90%). CONCLUSIONS Endoscopic cyanoacrylate injection for gastric varices does not limit the spread of bacteria. The endoscope accessory channel was the major source of bacteria. Most episodes of bacteremia were transient and uneventful.