Since the availability of screening tests for hepatitis C virus changes in the epidemiologic data have occurred, following the development of preventive and therapeutic measures and programs of large-scale screening. In the early 90's, transfusion was the major cause of contamination and the use of intravenous illicit drug was only in the second place. At this time, HCV infection, and this delay accounts for the high frequency of liver cirrhosis at diagnosis (20%). Since 1996 when screening by family physicians was promoted diagnosis is made more often because of the presence of risk factors in asymptomatic patients. This accounts for most of the observed changes in epidemiology of the infection: intravenous drug use is nowadays the most frequent source, whereas the proportion of transfused-borne cases has decreased. For the same reasons, hepatic lesions at diagnosis are less severe, and cirrhosis less frequent (10%), despite a similar delay between exposure and diagnosis. The current causes of hepatitis C amplify these changes: tranfusion-borne hepatitis have become an exception, nosocomial infections are much less frequent with the universal use of disposable devices and strict protocols for desinfecting instruments. On the contrary, the contamination by intravenous drug use does not seem to decrease and could account for 70% of new cases of seroconversion. The epidemiological surveillance has to be maintened inside networks, but requires simple indicators: the number of new cases. the mode of contamination, the severity of hepatic lesions at diagnosis. Preventive measures should be directed at intravenous drug users, and include campaigns in jails. As a conclusion, although the peak level of hepatitis C virus epidemic seems to be behind us, many infected people remain to be diagnosed and managed, and the risk exists in the next years to an increase in the incidence of hepatitis C virus related cirrhosis and hepatocellular carcinomas. (C) 2000 Editions scientifiques et medicales Elsevier SAS.