The aim of this study was to determine the outcome of venous conduits used in living donor liver transplantation (LDLT). We analyzed the portal vein complications in 66 LDLT recipients and 48 cadaveric reduced-size liver transplant (RLT) recipients performed from November 1989 through January 1995. Three different venous conduits were utilized in the LDLT recipients: Group 1, reconstructed vein from the living donor, n=18; Group 2, cadaveric cryopreserved iliac vein, n=37; and Group 3, cadaveric cryopreserved femoral vein, n=11. Overall, 47 percent of the patients were less than one year of age; the age distribution was not significantly different among the groups. The incidence of early thrombosis was significantly greater in LDLT Group 1, (33%) than any of the other groups (LDLT Group 2, 8%; LDLT Group 3, 9%; and RLT, 4%:P<0.0005 vs. reduced graft and < 0.03 vs. other LDLT groups). The incidence of late portal vein stenosis or thrombosis was significantly higher in the LDLT Group 2, (51%) than any of the other groups (LDLT 1, 16%; LDLT Group 3, 9%; RLT 4%;P<0.005 vs. cadaveric and < 0.02 vs. LDLT Group 1 and LDLT Group 3). Five year arterial graft and patient survival for patients who have experienced portal vein thrombosis or stenosis is 61% and 67%, respectively, versus 67% and 71% for those patients who have not experienced portal vein pathology, P=ns. Based on this experience, we recommend avoiding the use of cryopreserved iliac vein for portal vein reconstruction in liver transplantation. Every effort should be taken to eliminate the need for venous conduits in liver transplantation. If venous conduits must be utilized, cryopreserved femoral veins seem to provide superior patency rates. Careful clinical and ultrasonopraphic monitoring of patients at high risk for late venous thrombosis permits therapy with excellent graft and patient survival.