BACKGROUND/AIMS: A long-term study and follow-up of an uniform treatment strategy for esophageal varices was evaluated to plan future strategies more effectively. MATERIAL AND METHODS: Morbidity and mortality in 101 patients with liver cirrhosis followed up for 5 years in a single center after the first detection of oesophageal varices were studied. RESULTS: Definitive hemostasis of the first variceal bleeding episode could not be obtained in 31.5%, despite emergency endoscopic sclerotherapy in combination with vasopressin and/or balloon tamponade. Mortality related to this bleeding was 26% in this cohort of patients. Overall mortality from variceal bleeding during the 5 year period was 38%. Death occurred predominantly in patients with advanced liver disease, since the Child-Pugh score of patients who bled and survived was 8 +/- 2 as compared to 11 +/- 2 (p<0.01) in those who bled and died. If fundic varices were the origin of the hemorrhage, the mortality reached 78% within 6 months. Eighty-five percent of the fatalities were related to the liver disease. Besides variceal bleeding, infection was the predominant trigger leading to liver failure; occurring in 25%. The cause of non-liver related death was malignancy of esophagus, lung and pancreas. Only 13% of the patients with liver failure had undergone an orthotopic liver transplantation. CONCLUSION: Since variceal hemorrhage still carries a high mortality rate, primary prophylaxis for patients at high risk is indicated. The systematic, concomitant use of vasoactive drugs to sclerotherapy and the early performance of TIPS in case of failure of sclerotherapy have to be evaluated. Finally patients with an advanced liver disease and increased risk factors for variceal bleeding should be considered earlier for liver transplantation.