Introduction. Chronic hepatitis B and C (CHB and CHC) have a significant impact on morbidity and mortality. While the at-risk groups are known, the burden of these diseases is more difficult to assess. For the first time in France, we used data from the Medical Information Systems Program (PMSI) to assess their significance. Methods. All patients hospitalized with a primary diagnosis of CHB or CHC between 2004 and 2009 were included after linking all of their hospital stays. The analysis focused on sociodemographic characteristics, death, diagnoses of complications (cirrhosis, hepatocellular carcinoma), psychiatric disorders, and risk factors for severity (HIV co-infection, alcohol and psychoactive substance dependence). A Cox model analysis of mortality among patients with a single infection (HBV or HCV) identified factors associated with death during follow-up through December 31, 2009. Results. We identified 43,592 patients with HBV or HCV infection. Half had more than one hospitalization after the initial diagnosis of hepatitis, and 1,456 deaths occurred during follow-up. In univariate analysis, HCV infection significantly increased the risk of death compared to HBV, as did residence outside the Île-de-France region, age over 30 at diagnosis, complications, hospitalizations for psychiatric disorders or substance use disorders, and HIV co-infection. In multivariate analysis, all these factors remained significantly associated with death except for the type of hepatitis. Discussion and conclusion. This analysis was useful for assessing the burden associated with CH. Our results suggest that the prognosis does not differ by virus type after adjustment. While the interpretation of the results must remain cautious due to the limitations and biases associated with the PMSI, this initial attempt opens up possibilities for its use in the epidemiological surveillance of CH. (R.A.)