Data on the prevalence of chronic liver disease, derived from selected series of hospitalized patients or from mortality registers, underestimate the prevalence of chronic liver disease. The Dionysos Study is a cohort study that investigated for the first time the prevalence of chronic liver disease in a general population. All the citizens of two towns in northern Italy, Campogalliano and Cormons, aged 12 to 65 yr were contacted by letter. From March 1991 through March 1993, 6,917 of a total of 10,150 citizens were enrolled (compliance, 69%). The standardized protocol for each enrollee included (a) a color-illustrated food questionnaire on dietary habits and alcohol intake; (b) a detailed medical history, including questions on risk factors for chronic liver disease; (c) a physical examination; and (d) blood tests for AST, ALT, gamma-glutamyltranspeptidase, mean cell volume, platelet count and hepatitis B virus and hepatitis C virus markers. Signs suggestive of chronic liver disease were seen in 21.3% of the subjects, and who then underwent further liver function tests, upper abdominal ultrasonography and, when necessary, liver biopsy. Persistent signs of chronic liver disease were present in 17.5% of the subjects, including 1.1% with cirrhosis and 0.07% with hepatocellular carcinoma. The prevalence rates of hepatitis B virus and hepatitis C virus positivity (second-generation enzyme-linked immunosorbent assay) were 1.3% and 3.2%, respectively. Alcohol abuse was the etiological agent in 23%.(ABSTRACT TRUNCATED AT 250 WORDS)