This study comprised 100 persons with antibodies to hepatitis C virus (HCV), including 77 intravenous drug users (IVDUs). They were tested with serological HCV typing assays (Murex HCV serotyping 1-6 assay; Chiron RIBA HCV Serotyping SIA). Patients with a positive polymerase chain reaction (PCR) for HCV (n = 66) were tested with genotyping molecular assays (Inno-Lipa HCV II test; Sorin GEN-ETI-K HCV typing assay). Comparison of the results of these tests showed that (a) 92% of samples could be typed by one test at least; 44% could be typed by all four tests; 88% could be typed by one serological test at least and 66% by one molecular test at least; (b) 81% of the samples successfully tested with both serological tests gave comparable results; 95% of the samples successfully tested with both molecular tests gave comparable results; (c) when serological and molecular tests yielded different results, sequences in the 5'-non-coding (5' NC) or E1 regions always confirmed the results of the molecular tests; (d) in case of discrepancy between the results of the molecular tests the E1 region sequences confirmed the Sorin test results. It is concluded that the molecular tests compared gave similar results. The fact that the Murex serological test gave comparable results in more than 80% of cases indicates that it is an alternative to the molecular tests for routine diagnosis. However, comparison of the results of this test with those obtained in patients consulting a hepatology department showed that it gave the best results in a population of patients not exposed repeatedly to HCV.