PURPOSE: Risk factors for hepatitis C virus (HCV) infection have rarely been estimated using incident case-control studies in the "general" population. We undertook a case-control study of incident HCV infection to identify persistent modes of transmission in France. METHODS: Two types of case-patients were included: (1) repeat blood donors who seroconverted between 1998 and 2001 (with a last negative third-generation test reported from 1995 or after) and (2) seroconverters referred to hepatology departments in 2000 through 2001. For each case-patient, four age- and sex-matched controls were randomly selected from the population of occurrence. Data on risk factors were recorded for each case-patient's and matched control's referent exposure period (between last negative and first positive tests). RESULTS: Sixty-four case-patients and 227 controls were included. In univariate analysis, endoscopy (matched odds ratios [mORs] = 8.0; 95% confidence intervals [CI] = 2.3-27.2), general anesthesia (mOR = 5.6; 95% CI = 2.2-14.7), tattooing or body piercing (mOR = 8.8; 95% CI = 1.7-44.1), and intravenous (IV) drug use (p < 0.0001; mOR not defined) were associated with HCV seroconversion. In multivariate analysis, risk factors associated with HCV seroconversion were drug use (adjusted OR [aOR] = 109.0; 95% CI = 11.7-1015.8), digestive endoscopy (aOR = 5.7; CI = 1.4-23.8), and invasive radiology procedures (aOR = 11.6; CI = 1.7-78.5). CONCLUSIONS: The results showed the continuing major role of IV drug use and suggested that transmission related to invasive health care remained a potential source of new HCV infection between 1995 and 2001.