OBJECTIVE: To compare the transcatheter arterial chemoembolization (TACE) alone or plus radiofrequency ablation (RFA) in the treatment of single branch portal vein tumor thrombus(PVTT)in patients with hepatocellular carcinoma (HCC) so as to evaluate the safety, control rate, prognostic factors and overall survival. METHODS: From January 2004 to December 2007, 50 HCC patients (< 5 cm in diameter and 3 parenchymal lesions) with concurrent PVTT were enrolled and treated by TACE alone or TACE plus RFA randomly (TACE, n = 25; TACE-RFA, n = 25). In TACE group, the intra-hepatic lesions received TACE sequentially with RFA; in TACE-RFA group, PVTT and intra-hepatic lesions were treated with TACE sequentially with RFA separately. Strict follow-up was conducted by computed tomography and alpha-fetoprotein (AFP) assay. The survival time was analyzed by the Kaplan-Meier method and Cox regression analysis was performed to evaluate the prognostic factors. RESULTS: Of all 50 HCC patients with single branch PVTT with TACE or RFA, 47 patients (TACE, n = 24; TACE-RFA, n = 23) received all the scheduled procedures and completed the follow-up. Two patients (8.3%) in TACE group had liver dysfunction versus none in TACE-RFA group, 2 patients (8.7%) developed bile duct injury in TACE-RFA group related with the RFA procedure. The OR (overall response) for PVTT was 54.2% (complete response (CR) 8.3%, partial response (PR) 45.8%) in TACE group while 87.0% (CR 60.9%, PR 26.1%) in TACE-RFA group during the follow-up. From the definite diagnosis of HCC, the median survival was 8 months. And the 1-, 2- & 3-year survival rates were 33.3%, 12.5%, 8.3% in TACE group. And 26 months, 65.2%, 47.8%, 30.4% in TACE-RFA group respectively. The difference between two groups was significant. From the definite diagnosis of PVTT, the respective data were 7 months, 12.5% and 4.2%, 0 in TACE group versus 22 months, 52.2%, 34.8%, and 8.7% in TACE-RFA group with a significant P value. In multivariate analysis, only therapy (TACE or TACE-RFA) showed a protective value (hazard rate 0.430 vs 0.345, P < 0.05). Survival was not correlated with age, intra-hepatic tumor status, liver functions and AFP level for all patients. CONCLUSION: RFA is both safe and efficacious to prolong survival in the treatment of single branch PVTT plus TACE in selected HCC patients. It may provide rationales for further studies of evaluating the outcome of RFA plus other therapies in the treatment of HCC with single branch PVTT.