To test the theoretical possibility of 5'-UR mistyping between hepatitis C virus subtypes 1a and 1b, we combined a 5'-UR/Core line probe assay (LiPA) with a nested PCR system and retested 183 sera, previously genotyped as type 1a or 1b and originating mainly from Western Europe. Eight percent of these were found to be wrongly subtyped. Based on this method, 3 additional subtypes of type 1 were discovered (1d-1f). Randomly selected European type 2 sera (n = 18) were tested with a similar type 2 5'-UR/Core LiPA. They were unexpectedly found to belong to subtype 2c in the majority of cases. Among serum samples originating from South-East Asia, several additional genotypes (7a, 7c, 7d, and 9a) were detected which had 5'-UR sequence motifs indistinguishable from genotype 1. Based on 13,203 pairwise comparisons in the 340-bp NS5B region, classification into types, subtypes, and isolates was obtained in 99.8% of all cases by using the phylogenetic border value of 0.328 for subtypes/types and 0.127 for isolates/subtypes; and evidence for a 10th major type of HCV was provided. Combination of all available HCV sequence data from the 447-bp Core/E1 region and the NS5B 340-bp and 222-bp regions provided evidence for the existence of 10 types, including 50 subtypes. Previously, extensive studies involving genotypes 1a, 1b, 2a, and 2b indicated the importance of HCV subtyping in interferon treatment and progression of chronic liver disease. The herein described expansion in the number of HCV types and subtypes should help improve diagnosis, treatment and possibly prophylaxis of hepatitis C liver disease.