From January 1, 1978 to January 1, 1980 a controlled randomized trial comparing conservative treatment with prophylactic sclerotherapy of esophageal varices prior to hemorrhage was carried out. In all 71 patients liver cirrhosis was histologically confirmed. The two randomly assigned groups were comparable. Indications of endoscopic treatment were the existence of varices III-IV bearing erosions, varices II-IV without erosions but coagulation factors below 30%, or both. Six patients left the trial. In group Ia -- treatment by conservative means -- a high rate of variceal bleeding and death was observed. Comparing these results with those of group Ib treated by sclerotherapy, bleeding and death rates were found to be highly significantly lower. -- Thus the investigated criteria for predicting a recent variceal hemorrhage are confirmed. Prophylactic sclerotherapy in esophageal varices with erosions and/or poor coagulation reserve of the liver can largely prevent an esophageal hemorrhage from varices, and prolongs the life of these chronically ill patients.