In this study, we try to investigate the possible causes of failed TACE when it is served as a down-staging therapy for advanced hepatocellular carcinoma. Two hundred and seventy eight patients had diagnoses of unresectable hepatocellular carcinoma but had the possibility to accept a resection if they accepted the successful use of TACE as a down-staging therapy. The patients were divided into two groups based on the result of the down-staging therapy: those within the UCSF criteria (group 1, N = 166) and those outside of the criteria (group 2, N = 112). There was no significant difference between the two groups in their demographic characteristics or their liver function. There were more tumor targets in the failed group (2.5 ± 1.5) than in the successful group (2.1 ± 1.3) (P = 0.013); however, the difference in the total diameter of the targets per patient did not reach statistical significance (P = 0.321), after one to four TACE sessions, the final AFP level was significantly different between the two groups (P = 0.042); there were more patients with a post-treatment AFP level 400 ng/ml in the failed group(28.7% vs. 20.5%, P=0.042). After comparing the number of TACE sessions between two groups, no significant difference was found (P = 0.659). A logistic regression analysis revealed three independent factors that were associated with the failure of TACE when it was used as a down-staging therapy: the tumor target number, post-TACE AFP level and number of TACE sessions (P < 0.05). The current results indicate that post-TACE AFP ≥ 400 ng/ml and more tumor targets were the dropout factors influencing the failure of TACE when it was used as a down-staging therapy for advanced HCC patients who previously met the UCSF criteria.