The role of diet, except for alcohol drinking and aflatoxin contamination, in the etiology of hepatocellular carcinoma (HCC) is unclear. A hospital-based case-control study was conducted in Italy in 1999–2002, including 185 incident, histologically-confirmed cases of HCC. Controls were 412 subjects admitted to hospitals for acute, nonneoplastic diseases unrelated to diet. Dietary habits were assessed using a validated food-frequency questionnaire. Odds ratios (ORs) and the corresponding 95% confidence intervals (CI) were computed using unconditional multiple logistic regression, adjusting for hepatitis B (HBV) and hepatitis C (HCV) virus infection and alcohol drinking. Energy adjustment was carried out by means of the residual model. A significant inverse relation was found between intakes of milk and yoghurt (OR = 0.28; 95% CI: 0.13–0.61), white meats (OR = 0.44; 95% CI: 0.20–0.95), eggs (OR = 0.31; 95% CI: 0.14–0.69), and fruits (OR = 0.48; 95% CI: 0.22–1.05) and HCC risk. The favourable effect of high intakes of milk and yoghurt, white meats, eggs and fruits was consistent across strata of HBV and HCV infections. The present study supports the hypothesis of a role of diet in HCC aetiology. Dietary modifications may be indicated in subjects at high-risk for HCC. © 2006 Wiley-Liss, Inc.