In 1996 the prevalence, risk factors, and genotype distribution of hepatitis C virus (HCV) infection were assessed in the general population of a town in southern Italy. The sample was selected from the census by a systematic 1:4 sampling procedure. The participation rate was 96.6%. Among the 1,352 subjects enrolled, 195 (14.4%) tested reactive to antibody to HCV (anti-HCV) with enzyme immunoassay (EIA 3). When further tested with recombinant immunoblot assay (RIBA 3), 170 subjects (87.2%) tested positive, 23 subjects (11.8%) had indeterminate results, and 2 subjects (1%) tested negative. Thus, the overall anti-HCV EIA-positive RIBA-confirmed prevalence was 12.6% (170 of 1,352 subjects) and increased from 1.3% in subjects younger than 30 years to 33.1% in those > or =60 years of age. This latter age group accounted for 72.3% of all anti-HCV-positive subjects. Females tested positive more frequently than males (14.1% vs. 10.5%; P < .05). Alanine transaminase (ALT) concentrations were abnormal in only 4.1% (7/170) of anti-HCV EIA-positive RIBA-confirmed subjects. This suggests that ALT screening is not useful in the detection of anti-HCV-positive subjects in a general population. The results of multiple logistic regression analysis showed that an age of less than 45 years, the use of glass syringes, and dental therapy were all independent predictors of anti-HCV positivity. HCV RNA was detected by polymerase chain reaction in 75.9% of the 195 anti-HCV EIA-positive subjects: in 84.7% (144/170) of the RIBA-confirmed subjects; in 17.4% (4/23) tested as RIBA indeterminate; and in neither of the two subjects who tested RIBA negative. HCV type 1b was detected in 75 subjects (50.7%), type 2b in 1 subject (0.7%), type 2c in 66 subjects (44.6%), type 3a in 4 subjects (2.7%), and type 4 in two subjects (1.3%). These figures differ from those of Italian patients with chronic liver disease in whom genotype 2 is more rare. None of the individuals was infected with more than one genotype. The distribution of the two most common HCV viral types (1b and 2c) was not statistically different in terms of mean age, sex, or risk factors and suggests that they may have had a parallel spread in this community. These findings provide one of the highest overall anti-HCV prevalence rates in a general population with a likely cohort effect, i.e., decreased risk of infection along generations. These observations may indicate an epidemic or focus of hepatitis C that occurred several years earlier. The majority of anti-HCV-positive subjects in the oldest age group and with no clinical evidence suggests that HCV infection is a very prolonged and indolent disease.