Diagnostic testing for hepatitis C virus (HCV) infection currently is based on the presence of anti-HCV antibodies or a positive HCV RNA polymerase chain reaction (PCR) test. Although HCV RNA PCR is a sensitive and specific technique, widespread application is limited. Moreover, HCV RNA PCR is subject to false-positive reactions through contamination and is inherently difficult to standardize and quantitate. To overcome limitations of HCV RNA PCR, we produced both cDNA and riboprobes from a 241 nucleotide sequence of the 5' untranslated region of the HCV genome for slot hybridization. Hybridization was absent using normal human serum, horse serum, or hepatic cellular RNA from noninfected liver. Hybridization occurred predominantly with positive-stranded HCV RNA and was abolished by pretreatment with RNase A. Slot hybridization was performed on serum samples from 60 patients with chronic HCV infection and a positive HCV RNA PCR and 20 patients with liver diseases unrelated to HCV who had a negative HCV RNA PCR. Slot hybridization with cDNA and riboprobes showed concordance with HCV RNA PCR of 95 and 98.3%, respectively. There were no false-positive reactions in controls. The sensitivity of riboprobe hybridization was comparable to that of one stage HCV RNA PCR using 5' untranslated region primers. Riboprobe hybridization with the HCV H strain standard was positive in the dilution corresponding to 10(-6) chimpanzee infectious doses50/ml. The density of the hybridization signals correlated significantly with the mass of an RNA standard extracted from the liver of a patient with HCV infection. The relative quantities of HCV RNA in the sera of selected patients varied and were not correlated with the duration of disease or the histopathological stage. The highest relative quantities were associated with concurrent immunosuppression. We conclude that slot hybridization is a sensitive, specific alternative to HCV RNA PCR that can be directly quantitated using appropriate HCV RNA standards.