To assess pre-orthotopic liver transplantation (OLT) factors that could be evaluated pre-operatively or controlled post-operatively associated with hepatocellular carcinoma (HCC) recurrence and disease-free survival after liver transplantation (LT). Four hundred and twelve patients transplanted for HCC between 1988 and 1998 in 14 French centers, who survived the postoperative period were studied. Kaplan Meier estimates were calculated for 24 variables potentially associated with recurrence of HCC. Uni- and multivariate analyses were conducted to identify independent predictors of recurrence. Overall 5-year disease-free survival was 57.1%. By univariate analysis, variables associated with disease-free survival were: presence of cirrhosis (P = 0.001), etiology of liver disease (P = 0.03), alpha fetoprotein level (< 200, 200 to 2000, or > 2000; P < 0.0001), gamma-GT activity (N, N to 2N or > 2N; P = 0.02), the number of nodules (1, 2-3 or > or = 4; P = 0.02), maximal diameter of the largest nodule (< 3 cm, 3 to 5 cm or > 5 cm; P < 0.0001), the sum of the diameter of the nodules (< 3 cm, 3 to 5 cm, 5 to 10 cm or > 10 cm; P < 0.0001), bi-lobar location (P = 0.01), preoperative portal thrombosis (P < 0.0001), peri-operative treatment of the tumor (P = 0.002) and chemoembolization (P = 0.03), tumor differentiation (P = 0.01), initial type of calcineurin inhibitor (P = 0.003), the use of antilymphocyte antibodies (P = 0.02), rejection episodes (P = 0.003) and period of LT (P < 0.0001). By multivariate analysis, 6 variables were independently associated with HCC recurrence: maximal diameter of the largest nodule (P < 0.0001), time of LT (P < 0.0001), tumor differentiation (P < 0.0001), use of anti-lymphocyte antibody (ATG) or anti-CD3 antibody (OKT3) (P = 0.005), preoperative portal thrombosis (P = 0.06) and the number of nodules (P = 0.06). This study identifies immunosuppression, through the use of ATG or OKT3, as a predictive factor of tumor recurrence, and confirms the prognostic value of tumor differentiation.