One hundred and twelve consecutive Child Class A and B cirrhotic patients were included in a prospective controlled trial aimed at investigating the efficacy and safety of endoscopic sclerotherapy vs. distal splenorenal shunt in the elective treatment of hemorrhage from esophagogastric varices. Fifty-seven patients were randomly allocated to splenorenal shunt and 55 to endoscopic sclerotherapy. Since only 4 of the 55 patients assigned to endoscopic sclerotherapy had to be excluded after randomization and before treatment as compared to 14 of the 57 patients assigned to splenorenal shunt, it is suggested that the applicability of endoscopic sclerotherapy is greater than that of splenorenal shunt. One patient in each group died within 30 days of the procedure and two in the endoscopic sclerotherapy group were lost to follow-up just after discharge. Variceal rebleeding during follow-up occurred in 37.5% (18/48) of patients in the endoscopic sclerotherapy group and in 14.3% of those in the splenorenal shunt group (6/42) (p less than 0.02), whereas hepatic encephalopathy was more frequent in patients submitted to splenorenal shunt (10/42, 24%) than in those treated by endoscopic sclerotherapy (4/48, 8%) (p less than 0.05). The therapeutic modality was the only variable with independent predictive value for rebleeding during follow-up, whereas for hepatic encephalopathy, the therapeutic modality, and the presence of encephalopathy related to the bleeding episode each showed independent predictive value. Early and long-term mortality, did not differ between the two therapeutic groups, being the 2-year survival was 71% for splenorenal shunt and 68% for endoscopic sclerotherapy.(ABSTRACT TRUNCATED AT 250 WORDS)