This prospective study was undertaken to assess the prevalence of Dupuytren's contracture (DC) and its relationship with possible causes, especially alcohol consumption and chronic liver disease. Four hundred thirty-two consecutively hospitalized patients were examined for evidence of DC. They were divided into five groups based on the following clinical, biologic, and histologic criteria: alcoholic cirrhosis (89 patients), noncirrhotic alcoholic liver disease (55 patients), chronic alcoholism without liver disease (46 patients), nonalcoholic chronic liver disease (68 patients), and a control group (174 patients). The prevalence of DC in these five groups of patients was 32.5%, 22%, 28%, 6%, and 12%, respectively; the prevalence of DC was higher in patients with cirrhotic or noncirrhotic alcoholic liver disease (25.5%) than it was in patients with nonalcoholic liver disease (6%), but it was not significantly different in alcoholic patients with or without liver disease. The relationship between DC and age, sex, manual labor, previous hand injuries, diabetes mellitus, alcohol consumption, and cigarette smoking was assessed by univariate and logistic regression methods. Nine variables were significantly different in patients with or without DC: age, sex, manual labor, previous hand injuries, diabetes mellitus, daily alcohol consumption, duration of alcohol consumption, total alcohol consumption, and duration of cigarette smoking. In our patients, variables that could explain DC were, in decreasing order, age, total alcohol consumption, sex (male), and previous hand injuries. In alcoholic patients, these variables were age and previous hand injuries; in nonalcoholic patients, these variables were age and cigarette smoking. These results emphasize the high prevalence of DC in alcoholic patients and the absence of a correlation between DC and chronic liver disease. Age and alcohol consumption are the best explanatory variables of DC in hospitalized patients.