BACKGROUND: The prognosis of patients with hepatocellular carcinoma (HCC) undergoing transarterial chemoembolization (TACE) is variable, despite a myriad of prognostic markers. We compared and integrated the established prognostic models, HAP and ART scores, for their accuracy of overall survival (OS) prediction. RESULTS: In both training and validation sets, HAP and ART scores emerged as independent predictors of OS (p<0.01) with HAP achieving better prognostic accuracy (c-index: 0.68) over ART (0.57). We tested both scores in combination to evaluate their combined ability to predict OS. Subgroup analysis of BCLC-C patients revealed favorable HAP stage (p<0.001) and radiological response after initial TACE (p<0.001) as positive prognostic factors. PATIENTS AND METHODS: Prognostic scores were studied using multivariable Cox regression and c-index analysis in 83 subjects with Barcelona Clinic Liver Cancer (BCLC) A/B stage from UK and Italy (training set), and 660 from Korea and Japan (validation set), all treated with conventional TACE. Scores were further validated in an separate analysis of patients with BCLC-C stage disease (n=63) receiving initial TACE. CONCLUSION: ART and HAP scores are validated indices in patients with intermediate stage HCC undergoing TACE. The HAP score is best suited for screening patients prior to initial TACE, whilst sequential ART assessment improves early detection of chemoembolization failure. BCLC-C patients with low HAP stage may be a subgroup where TACE should be explored in clinical studies.