To compare efficacy of transarterial chemoembolization with and without radiation therapy (RT) versus sorafenib for advanced hepatocellular carcinoma (HCC) with portal vein tumor thrombosis (PVTT).This single-center retrospective study involved 557 patients with HCC with PVTT who initially received chemoembolization (1997–2002; n = 295), chemoembolization and RT (2003–2008; n = 196), or sorafenib (2009–2012; n = 66) according to eligibility criteria among an initial population of 617. The three groups were divided into three pairs (chemoembolization vs chemoembolization/RT, chemoembolization vs sorafenib, and chemoembolization/RT vs sorafenib), and time to progression (TTP) and overall survival (OS) were compared by propensity-score analyses.The chemoembolization/RT group had longer median TTP and OS than the chemoembolization-alone and sorafenib groups (P < .001). Multivariate Cox analysis revealed that chemoembolization/RT treatment was an independent predictor of favorable TTP and OS. In the matched cohort, median TTP and OS were significantly longer in the chemoembolization/RT group than the chemoembolization-alone group (102 pairs; TTP, 8.7 mo vs 3.6 mo [P < .001]; OS, 11.4 mo vs 7.4 mo [P = .023]) or the sorafenib group (30 pairs; TTP, 5.1 mo vs 1.6 mo [P < .001]; OS, 8.2 mo vs 3.2 mo [P < .001]), in agreement with the inverse probability of treatment weighted (IPTW) outcomes. In matching analyses, the chemoembolization-alone group had longer median TTP and OS than the sorafenib group (46 pairs; TTP, 3.4 mo vs 1.8 mo [P < .001]; OS, 5.9 mo vs 4.4 mo [P = .003]). There was no significant difference in terms of OS with the IPTW approach (P = .108), but there was one in terms of TTP (P < .001).Within the limitation of a retrospective study, the present data indicate that transarterial chemoembolization combined with RT could be considered as an alternative to the standard sorafenib in the treatment of patients with advanced-stage HCC with PVTT.