significantly associated with overall survival (OS) on univariate analysis and therefore included in multivariate analysis. Only Child– Pugh stage and HAP-Score were independent prognostic factors. The median OS for HAP A/B/C/D was 26/18/14/10 months (p < 0.0001). Conclusions: We could validate the HAP-score in an independent cohort of HCC patients treated with TACE. The HAP-score discriminates four different prognostic subgroups. Its clinical usefulness is limited by only 2 treatment options being available for this group of patients (TACE and drug treatment). A simplified score could be a useful tool for the selection of HCC patients for TACE.