THE ORIGINAL operation of portacaval shunt was first successfully carried out by von Eck in 1877.1 Since that time, sporadic operations have been performed,2 but results were discouraging and the operative death rate prohibitive. Today this operation may safely be recommended for properly selected patients.3 Advances in anesthesiology have contributed much to the safety of this procedure, and adequate administration of blood and fluid, preoperatively, postoperatively and more particularly during the operative procedure, have further secured that safety. Proficiency in the execution of anastomoses between delicate and friable veins is essential, and the surgeon must master these skills in the experimental laboratory before attempting to perform the procedures in human beings. The two cases presented in this report are illustrative of some of the technical difficulties which may be encountered in attempting to perform a portacaval shunt. REPORT OF CASES Case 1. —E. B. (St. Lukes, no.