A recently introduced enzyme immunoassay procedure for antibodies against the hepatitis-C virus (HCV) was used to test samples from 185 cases with hepatocellular carcinoma (HCC) and 432 hospital controls. The anti-HCV results were examined in conjunction with previously reported data from this study concerning hepatitis-B virus (HBV) serology, hepatitis-D virus (HDV) antibodies, presence of cirrhosis and tobacco smoking. There was evidence for interaction between HBV and HCV in the causation of HCC: as previously reported, the rate ratio (RR) linking the presence of anti-HCV to HCC among subjects positive for hepatitis-B surface antigen (HBsAg) was substantially higher than the corresponding RR among those negative for this marker; furthermore, among HCC patients positive for HBsAg, a high proportion (33/61) of those who were positive for hepatitis-Be antigen (HBeAg) or its antibody were positive for anti-HCV, whereas among HBsAg-positive controls who were also positive for HBeAg or its antibody, none was positive for anti-HCV (0/18; p less than 10(-4)). The anti-HCV-related RR for HCC was also higher among HCC patients with cirrhosis than among those without evidence of co-existing cirrhosis (RR 11.4 vs. 4.4; p = 0.06). In addition, there was some evidence of interaction between tobacco smoking and HCV in the origin of HCC; after controlling for age, sex and HBsAg status, the RR for subjects positive for anti-HCV was 6.8 among smokers but only 3.2 among non-smokers (p = 0.26). By contrast, there was no suggestion of an interaction between anti-HCV and anti-HDV, in agreement with the presumed minimal role, if any, of HDV in HCC etiology. These results support the notion that HCV is involved in the etiology of HCC by advancing, through a chronic liver disease process, carcinogenesis initiated by other factors.