BACKGROUND Portasystemic shunts, especially total shunts, are effective tools for reducing portal pressure and controlling variceal bleeding but lead to high risk of encephalopathy and accelerating liver failure. The purpose of this study is to evaluate the clinical effects of small-diameter expanded polytetrafluoroethylene (ePTFE) H-graft portacaval shunts in the treatment of portal hypertension. METHODS Thirty-one patients with portal hypertension were treated with ePTFE small-diameter H-graft portacaval shunts from December 1995 to April 2002. Twenty-one had externally ringed grafts and 10 had non-ringed grafts; 20 had 10 mm diameter grafts and 11 had 8 mm grafts. The left gastric artery and coronary vein were ligated in 22 patients. Additionally, 6 patients underwent pericardial devascularization, and splenectomies were performed on 30 patients. RESULTS An average decrease of free portal pressure (FPP) from (32.13 +/- 4.86) cmH2O before shunting to (12.55 +/- 5.57) cmH2O after shunting was observed. Portal blood flow was reduced by 1/3 compared with the levels measured before shunting. Twenty-eight patients survived after the operation, and no upper gastrointestinal rebleeding occurred in the follow-up period (40.2 months on average). We lost contact with one patient. Color Doppler ultrasonography and/or portography revealed the shunts to be patent in 28 cases and occluded in 2 (6.4%) cases. Encephalopathy developed in 4 patients (12.9%). CONCLUSIONS Small-diameter ePTFE H-graft portacaval shunts can effectively reduce portal pressure. Moreover, the majority of the hepatopetal flow from the portal vein can be adequately maintained. The reinforced shunts may achieve a higher rate of patency. Morbidity from encephalopathy was less frequent than in patients receiving total shunts. Small-diameter H-graft portacaval shunts are also effective in preventing recurrent variceal bleeding.