To compare immunohistochemical and molecular methods for the detection of hepatitis C virus (HCV) infection in archival liver biopsies we analyzed formalin-fixed and paraffin-embedded liver specimens of 10 patients with serologically confirmed HCV infection. Methods employed included indirect FITC-immunofluorescence, reverse-transcriptase polymerase chain reaction (RT-PCR) using extracted RNA and Southern blotting with chemiluminescence-based detection, non-radioactive in situ hybridization (ISH) with digoxigenin-labeled oligo- and cRNA probes, direct in situ RT-PCR with incorporation of labeled nucleotides into PCR-products, and indirect in situ RT-PCR using subsequent ISH for the visualization of intracellular PCR-products. Our results indicate that: (1) using the histological criteria described by Lefkowitch et al. (Gastroenterology 1993; 104-595) together with clinical data, most chronic HCV infections can be diagnosed by conventional histology, if liver biopsies are representative; (2) the commercially available mAB TORDJI-22 appears to cross-react with non-HCV epitopes; (3) molecular methods performed on routinely fixed and processed liver biopsies frequently yield false negative results due to sampling problems, low viral copy number and RNA degradation in infected cells; (4) analysis of HCV-RNA by RT-PCR of extracted total RNA is more sensitive than indirect in situ RT-PCR or ISH; and (5) direct in situ RT-PCR is not reliable despite the use of modifications such as DNase pretreatment and hot-start procedures. Further studies are required to define both optimal methods for sample processing and improvements of protocols, in order to increase detection sensitivity and specificity of HCV infection by immunohistochemical and molecular methods.