If a portacaval shunt is to be considered satisfactory treatment in the prevention of hemorrhage from gastroesophageal varices, it should be offered a cirrhotic patient with intrahepatic obstruction of portal blood flow and a functioning coronary vein visualized by splenoportography. The value of the prophylactic shunt would depend upon the significance of the coronary vein to bleeding varices. The importance of such a consideration was suggested early in our experience. Studies on three cirrhotic patients indicated that a clinical significance may be attached to the first collateral channels, or branches of the splenic vein, filling with dye during these venographic studies, particularly, the prompt opacification of the coronary vein. In one patient (Fig 1A) with frequent hemorrhages from varices, a large coronary vein represented the only venous channel providing a route for portosomatic blood flow. In a second instance (Fig 1B), the only vein so functioning was a