In view of the proven efficacy of endoscopic sclerotherapy and the even improving results of liver transplantation, the present role of porto-systemic shunt should be reconsidered. From 1986 (when our liver transplant program began), to March 1994, 59 cirrhotic patients (males = 40, females = 19, mean age 53.17 +/- 12.04) underwent a porto-systemic shunt, 22 under emergency conditions and 37 in an elective setting. Patients were subdivided according to age, emergency or elective surgery, type of operation, and liver function. In the emergency procedures previous sclerotherapy and time between admission and surgery were also considered in the assessment. Mean follow-up was 46.49 +/- 31.48 months. Overall 5-year actuarial survival was 62.5%. In the emergency porto-systemic shunts the worst short-term results were obtained in patients over 55 years of age (p < 0.05) and when operations were performed within the first 24 hours after admission (p < 0.005). Long-term survival was not significantly influenced by the variables considered although patients over 55 years of age and patients with reduced liver function (Child B and C) seemed to have a more dismal outcome. Those patients under 55 years of age, with no portal thrombosis, considered as potential liver transplant candidates, had a better short-term survival rate (p < 0.05) than that of the rest of the patient population studied, mainly because of the better outcome after emergency surgery. Our data confirm the efficacy of porto-systemic shunt procedures in preserving the patient from variceal bleeding. They have a definite role in the complex treatment strategy of portal hypertension, and they must not be considered only a rescue procedure. However, liver transplantation remains the best option to resolve both portal hypertension and the underlying liver disease.