In patients with chronic hepatitis C, determination of hepatitis C virus (HCV) genotype could be routinely run in the future to tailor treatment schedules. The suitabilities of two versions of a serological, so-called serotyping assay (Murex HCV Serotyping Assay version 1-3 [SA1-3] and Murex HCV Serotyping Assay version 1-6 [SA1-6]; Murex Diagnostics Ltd.), based on the detection of genotype-specific antibodies directed to epitopes encoded by the NS4 region of the genome, for the routine determination of HCV genotypes were studied. The results were compared with those of a molecular biology-based genotyping method (HCV Line Probe Assay [INNO-LiPA HCV]; Innogenetics S.A.), based on hybridization of PCR products onto genotype-specific probes designed in the 5' noncoding region of the genome, obtained with pretreatment serum samples from 88 patients with chronic hepatitis C eligible for interferon therapy. Definitive genotyping was performed by sequence analysis of three regions of the viral genome in all samples with discrepant typing results found among at least two of the three assays studied. In all instances, sequence analysis confirmed the result of the INNO-LiPA HCV test. The sensitivity of SA1-3 was 75% relative to the results obtained by the genotyping assay. The results were concordant with those of genotyping for 92% of the samples typeable by SA1-3. The sensitivity of SA1-6 was 89% relative to the results obtained by the genotyping assay. The results were concordant with those of genotyping for 94% of the samples typeable by SA1-6. Overall, SA1-6 had increased sensitivity relative to SA1-3 but remained less sensitive than the genotyping assay on the basis of PCR amplification of HCV RNA. Cross-reactivities between different HCV genotypes could be responsible for the mistyping of 8 (SA1-3) and 6% (SA1-6) of the samples. Subtyping of 1a and 1b is still not possible with the existing peptides, but discriminating between subtypes may not be necessary for routine use.