Between November 1996 and April 1998, 260 consecutive patients seen in our department with histologically proven chronic hepatitis C were studied prospectively. At the time of the study, none of the patients had been treated for hepatitis C or consumed alcohol at more than 50 g/day. The baseline characteristics of the patients are shown in Table 1. There were 150 men and 110 women, with a mean age of 42.9 ± 12.2 years. The route of HCV transmission was blood transfusion in 78 cases (30.0%), intravenous drug use in 80 cases (30.8%) and unidentified in 102 cases (39.2%). The body mass index was determined in 181 patients. Both serum anti-HCV antibodies (second- and third-generation enzyme-linked immunoabsorbent assays, Ortho Diagnostic Systems, Raritan, NJ, USA) and serum HCV RNA (Amplicor HCV, Roche Molecular Systems, Pleasanton, CA, USA) were present in all patients. Serum HBsAg and anti-human immunodeficiency virus antibodies were absent. The HCV genotype was determined in 255 patients (INNO-LIPA HCV II, Innogenetics, Ghent, Belgium). The HCV RNA load was measured at the time of liver biopsy using a second-generation bDNA assay (Quantiplex HCV-RNA 2.0, Chiron Diagnostics, Emeryville, CA, USA) in 180 patients for whom frozen serum was available. All liver biopsies were fixed in formalin, embedded in paraffin and routinely processed for histological analysis. The grade of histological activity and the stage of histological fibrosis were assessed using the validated METAVIR scoring system, as follows: A1–A3 for the degree of necro-inflammatory activity (A1, mild activity; A3, marked activity) and F0–F4 for the degree of fibrosis (F0, absence of fibrosis; F4, cirrhosis).14, 15 Steatosis, graded as absent, mild (< 10% of hepatocytes), moderate (10–30% of hepatocytes) or marked (> 30% of hepatocytes), and histological lesions consistent with alcohol-induced liver disease (i.e. Mallory bodies and inflammatory infiltration predominantly composed of neutrophils) were recorded. In addition, a semi-quantitative histological grading in the range 0–3 (0, absent; 1, mild; 2, moderate; 3, marked) was used for the histological assessment of liver iron accumulation on Perls' staining. Alcohol intake was estimated on the basis of beverage-specific quantity frequency measures validated by the 1988 United States National Health Interview Survey.16 A questionnaire was given to each patient on the day of liver biopsy to determine the alcohol intake (number of drinks consumed per day), separately for beer, wine and spirits, on weekdays and at weekends, during the 6 months preceding evaluation. It was checked that, during this period, the alcohol intake was stable and was not influenced by the diagnosis of chronic hepatitis C. According to the level of alcohol intake, the patients were classified into four groups. Intake was considered to be minimal (1–20 g/day) in 96 patients (36.9%), mild (21–30 g/day) in 34 (13.1%) and moderate (31–50 g/day) in 37 (14.2%). Ninety-three patients (35.8%) denied alcohol intake and none of the patients consumed more than 50 g/day.