Background There has been no consensus as to which system, either the Cancer of the Liver Italian Program (CLIP) or the Japan Integrated Staging (JIS) system, is suitable to predict the prognosis of hepatocellular carcinoma (HCC) patients who underwent transcatheter arterial chemoembolization (TACE) as initial therapy. Purpose To retrospectively compare the usefulness of CLIP and JIS in predicting and stratifying the prognosis of HCC patients treated by TACE. Material and Methods Between 1995 and 2005, consecutive 728 patients with untreated HCC who underwent TACE in our institute were selected for this study. The survival rate and its prognostic factors were assessed by multivariate analysis. Patients were stratified according to the two systems, and their survival rates between the scores were compared. Results The mean follow-up period was 1689 days. The 1-year, 3-year, 5-year, and 10-year survival rates were 83.1%, 55.1%, 34.7%, and 12.8%, respectively. Both systems stratified the prognosis of patients well, but was slightly better in CLIP as compared to in JIS. As for multivariate factor analysis, less severe Child-Pugh classification ( P < 0.001), simple tumor morphology ( P < 0.001), absence of portal vein invasion ( P < 0.001), and lower alpha-fetoprotein (AFP) level ( P < 0.001) were suggested to be independent indicators for favorable survival rate. All of these independent factors were included in CLIP, whereas JIS lacked AFP level. Furthermore, the likelihood χ-test value was higher, and the Akaike information criterion value was lower for CLIP than for JIS. Conclusion CLIP is more suitable than JIS for predicting prognosis of patients with HCC who would undergo TACE in a Japanese population.