INTRODUCTION: After curatively intended surgery for colorectal liver metastases, liver recurrences occur in more than 60% of patients, despite the administration of adjuvant systemic chemotherapy. The aim of this study was to assess the benefit of combined adjuvant hepatic arterial infusion (HAI) and intravenous (IV) 5-FU compared with standard modern adjuvant IV chemotherapy in patients at high risk of hepatic recurrence. PATIENTS AND METHODS: From January 2000 to December 2009, 98 patients, who had undergone curative resection of at least 4 colorectal liver metastases, were selected from a prospective database. Among them, 44 (45%) had received postoperative HAI combined with systemic 5-FU (HAI group) and 54 (55%) had received "modern" systemic chemotherapy (IV group). RESULTS: The 2 groups were similar in terms of age, sex, the stage of the primary, and the administration of preoperative chemotherapy. The median number of HAI cycles received per patient was 7 [range, 1-12]. Twenty-nine patients (66%) had received at least 6 cycles of HAI oxaliplatin, and 22 patients (50%) had received the full planned treatment. For the remaining 22 patients (50%), HAI chemotherapy had been discontinued because of toxicity (n = 8), HAI catheter dysfunction (n = 6), an early recurrence (n = 6), and patient's refusal (n = 2). After a median follow-up of 60 months (51-81 months), 3-year overall survival was slightly higher in the HAI group (75% vs 62%, P = 0.17). Three-year disease-free survival was significantly longer in patients in the HAI group than those in the IV group (33% vs 5%, P < 0.0001). In the multivariate analysis, adjuvant HAI chemotherapy and an R0 resection margin status were the only independent predictive factors for prolonged disease-free survival. CONCLUSIONS: Postoperative HAI oxaliplatin combined with systemic chemotherapy after curatively intended surgery of colorectal liver metastases is feasible and may significantly improve disease-free survival of patients at high risk of hepatic recurrence compared with adjuvant modern systemic chemotherapy alone. These results should be confirmed in a randomized study.