To determine risk factors for hepatocellular carcinoma (HCC) in Japan, we studied 575 patients with liver cirrhosis (LC) who presented to our out-patient clinics in a month; 217 patients with habitual drinking with or without viral infections (AL-LC), 235 with hepatitis C virus (HCV) infection, 40 with hepatitis B virus (HB) infection, and 82 with liver diseases from other aetiologies such as primary biliary cirrhosis (PBC). HCC was found in 28% of AL-LC, 27% of HCV-LC, and 18% of HB-LC. HCV antibody examined by the second-generation HCV antibody assay (HCV II) was positive in 49% of AL-LC without HCC, and 84% of AL-LC with HCC. HCV was significantly associated with the development of HCC with an odds ratio of 1.90 by multivariate analysis, whereas alcohol alone did not become a risk factor, with an odds ratio of 0.65. When both variables were combined (HCV+AL), however, the odds ratio was increased twice to 3.65. One-hundred and thirteen patients who had a history of blood transfusion more than 10 years ago were selected to assess the interaction between alcohol and HCV in hepatocarcinogenesis. A year-adjusted disease occurrence rate calculated by the Kaplan-Meyer method showed that HCV+AL had a significantly higher disease occurrence rate than HCV alone. These results suggest that although alcohol alone may not become a risk factor for HCC, it may potentiate the development of HCC caused by HCV. HCV+AL is considered to be the highest risk group; strict abstinence is necessary for these patients.