After a large uncontrolled experience, in April 1976, a randomized clinical trial between side-to-side portacaval anastomosis and a mesentericocaval shunt with internal jugular vein interposition was initiated. Up to April 1979, 23 patients were operated upon with a mesentericocaval shunt and 26 with a portacaval anastomosis. With a mean follow-up time of 23 months, no statistically significant differences in terms of operative mortality, hepatic encephalopathy and long term survival were found between the two operations. Therefore, a mesentericocaval shunt does not appear to offer effective advantages compared with the portacaval anastomosis, and it should be performed in particular instances in which, for anatomic reasons, a portacaval anastomosis cannot be performed.