Aims: The aim of this study was to evaluate tbe efficiency of fibrin glue (FG) as an injectate agent for endoscopic ernbolization (EE) of esopbageal varices (EV). Metbods: From Marcb 1995 to June 1997, twenty seven consecutive patients (18 males and 9 females; mean age, 46.1±2.4 yr) were selected for a study on EE of EV using FG, at !he Department of Surgery, Hospital "St. Treime". The origin of portal hypertension was postbepatic (HBV and/or HCV) liver cirrhosis. Severity of liver disease was classified according to !he Child-Pugh classification: class NB/C-4/14/9, !he mean score on adm.ission was 8.96±0.4. Prophylactic EE (lgr) was performed in 15 pts with high-risk EV(F2 or F3, RC sign -t+ or +++) according selection criteria JRSPH(l991)and EE for prophylactic of recurrent varicea! bleeding in 12 pts. An absolute criteria for exclusion was: pts who had already been treated endoscopically for EV. FG was prepared ex tempore: fibrinogen(30mg/ml), thrombin ( I 50300units/ml), CaCl2(40mM/ml), aprotinin (3.000 KIU/ml). EE was carried out with a flexible endoscope and an original or commercial injectors by intravariceal or combined injections of FG. EE was repeated once a week until all EV has been eradicated. Results: Characteristics of EE with FG were: total sessions in botb gr.-33, total volume of sclerosant-6.6±-0.5(3-11 )ml, volume of FG per single varix 3. l ±-0.2(2-6)ml. Only minor complications (fever, chest pains) were observed in 4(14.8%). Ulcer formation, esophageal stenosis and perforation did not occur. The therapeutic end point (Cw-Th,RC-) in botb groups was achieved în all patients with a Jower mean number of sessions I .2(range 1-3). Cbanges in the selected hemostasis parameters were not essential, clinically there were no detectable thrombotic or heroorrbagic complications. Mean duration of follow-up was 25.6±2.4(1-44)months, during this period 9 pts died (6-bepatic failure and 3-bleeding from acute gastric and duodenal lesions). Conclusions: Tb.is study shows !bat FG is an efficient and safe agent for EE of EV. However, studies in !ager series and prospective randomized trials comparing FG witb otber sclerosant agents (inc. thrombin) are still needed before a final conclusion can be reached.