The results of liver transplantation for advanced hepatocellular carcinoma have been disappointing because of high recurrence rates, leading to low long-term survival; this indication remains controversial in an era of organ shortage. To continue to offer this treatment possibility, efforts were made to reduce recurrence and improve survival. The first approach is to maintain a strict selection policy excluding patients with extraheptic disease. The second approach is to test the efficacy of perioperative adjuvant therapies in patients selected for transplantation. The reasons for recurrence include: (1) undetected preoperative micrometastases, (2) intraoperative dissemination by surgical manipulation, and (3) acceleration of tumor growth by immunosuppression. Preoperative treatment, which may include systemic chemotherapy or arterial chemoembolization, seems necessary to limit tumor progression during the waiting period. Systemic chemotherapy has been tested pre-, intra-, and postoperatively. It is considered essential postoperatively and should be used as soon as possible after surgery (i.e., in the 1st postoperative week). Several teams have performed pilot studies that included rather limited numbers of patients, and the results were compared with those for historic controls. Published results show that chemoembolization creates tumor necrosis in most instances. This is efficient in limiting tumor progression but the effect on recurrence and survival is unknown. All authors who used postoperative chemotherapy reported improved survival over controls, with 50%-60% 3-year survival and up to 50% 5-year survival. However, recent results suggest that late recurrence may occur. Chemotherapy was usually well tolerated, although leukopenia, sometimes severe, was observed in most patients. The use of granulocyte colony-stimulating factor seems useful to overcome this problem. Perioperative adjuvant treatments seem to prolong survival in patients undergoing liver transplantation for advanced hepatocellular carcinoma, but delayed recurrence remains possible. Further studies are necessary; these should ideally be multicentric, prospective, and randomized.