Recent studies have demonstrated that increased resistance to portal inflow is not solely responsible for the development of portal hypertension. Increased splanchnic flow has been attributed to a combination of factors, including elevated circulating levels of vasodilators and diminished sensitivity of the splanchnic vasculature to endogenous vasoconstrictors. In selected animal models of portal hypertension, increased splanchnic flow accounts for approximately 40% of the observed elevations in total portal venous pressure. Improved understanding of the pathophysiologic factors responsible for the development of portal hypertension has led to pharmacologic efforts to decrease portal pressure. Current limitations include lack of drug selectivity and specificity and inability to predict and monitor patient responses. Primary treatment options include selective portosystemic shunts, endoscopic sclerotherapy (ES), and orthotopic liver transplantation. ES is more effective in preventing recurrent variceal hemorrhage than medical treatment but is less effective than shunt surgery. In selected studies, ES better maintains hepatic function and may prolong survival compared to primary shunt surgery. ES failures occur in nearly 33% of patients, but "salvage shunts" in these patients appear to be reasonably safe and quite effective in preventing recurrent hemorrhage. Selective shunts are favored because they appear to confer a better quality of life (but not improved longevity) than conventional shunts. Liver transplantation is preferred for patients with end-stage liver disease in whom the predicted mortality of conventional surgery outweighs the survival benefit.