AIM: To prospectively evaluate the short and long term clinical impact of selective transarterial chemoembolization (TACE) on liver function in patients with hepatocellular carcinoma (HCC). To assess side effects in relation to treatments. To analyze the overall survival and HCC progression free survival probability. METHODS: One hundred and seventeen cirrhotic patients with HCC were enrolled. Baseline liver function included Child-Pugh score and serum levels of alanineaminotransferase (ALT), prothrombin time (PT) and bilirubin. According to Cancer Liver Italian Program (CLIP) and Barcelona Clinic Liver Cancer (BCLC) staging systems, 71 patients were eligible for TACE; 32 had previously received treatment for HCC. No significant differences in liver function were observed between previously treated and not treated patients. TACE was performed by selective catheterization of the arteries nourishing the lesions. While hospitalized, pat ients underwent cl inical , hematologic and ultrasonographic assessments. One month after TACE a CT scan was performed to assess tumor response. A second TACE was performed on demand. Liver function tests were checked in all patients every four months. RESULTS: After first TACE, the mean Child-Pugh score increased from a mean baseline 5.62 ± 1.12 to 6.11 ± 1.57 at discharge time ( P < 0.0001), decreasing after four months to 5.81 ± 0.73 (not significant). ALT, PT and bilirubin significantly ( P < 0.0001) increased 24 h after TACE and progressively decreased until discharge. After the second TACE, variations in Child-Pugh score, ALT, PT and bilirubin were comparable to that described after the first TACE. No major complications were observed. The mean follow-up was 14.7 ± 6.3 mo (median: 16 mo). Only one patient died. No other patient experienced important long term worsening of clinical status. The overall survival probability at twenty-four months was 98.18% with a correspondent HCC progression free survival probability of 69%. CONCLUSION: Selec t ive TACE may produce significant, but transitory increases in ALT values, with no major impact on liver function and Child-Pugh score. Preservation of liver function is achievable also in patients previously treated with other therapeutic modalities and in patients undergoing multiple TACE cycles. Liver function can remain stable in the long-term, with optimal medium term survival. This result can be achieved through rigorous patient selection on the basis of tumour characteristics and clinical conditions.