In patients with portal hypertension, the splenorenal shunt is one of the most common portosys- temic collaterals. Spontaneous hepatic encephalopathy frequently occurs in these patients, due to neurotoxic substances that bypass the liver and act on the central nervous system directly. Therefore it is important to assess portal hemodynamics in portal hypertensive patient with porto-systemic shunt. We examined the portal blood flow by the Doppler flowmetry system in 9 patients with liver cirrhosis and 3 patients with hepatocellular carcinoma associated with liver cirrhosis who had large spontaneous splenorenal shunts. Of the 12 subjects, 7 patients had a history of spontaneous hepatic encephalopathy and 5 patients did not. One of the 5 patients with a history of spontaneous hepatic encephalopathy and 4 of the 5 patients without such a history had mild esophageal varices. We could assess the direction of splenic venous flow in 6 patients with a history of spontaneous hepatic encephalopathy and they all showed reversed flow. Only one patient among the 5 patients without such a history showed reversed splenic venous flow. The portal vein diameter was signifi- cant)y smaller in patients with a history of spontaneous hepatic encephalopathy (0.88+0.27 cm) than in patients without this history (1.32+0.08 cm)(p<0.01j. We conc1uded that the direction of splenic venous flow and the diameter of the portal vein in portal hypertensive patients with a large splenorenal shunt may be important factors in the development of spontaneous hepatic encephalopathy.