We compared a single-enzyme, combined reverse transcription-PCR (RT-PCR; AMPLICOR HCV Test; Roche Molecular Systems, Branchburg, N.J.) with an independent, two-enzyme, standard RT-PCR (SRT-PCR) assay for the detection of hepatitis C virus (HCV) RNA in serum and plasma. Test samples included a proficiency testing panel consisting of 10 undiluted plasma samples, three separate dilution series, and sera from 99 patients with chronic liver disease. The quantity of HCV RNA in each patient serum sample was determined by a branched DNA (bDNA) signal amplification assay (Quantiplex HCV-RNA assay; Chiron, Emeryville, Calif.). There was complete concordance between the results of the RT-PCR assays with the 10 undiluted plasma samples used for proficiency testing (3 positive and 7 negative samples). However, the analytical sensitivity of SRT-PCR was 4- to 10-fold greater than that of the AMPLICOR test in the dilution series. HCV RNA was detected in 44, 45, and 40 of the patient serum samples, by SRT-PCR, the AMPLICOR test, and the bDNA assay, respectively. There was 97% agreement between the results of the RT-PCR assays, with only three discrepancies. Review of the patients' medical records resolved all three discrepancies in favor of the AMPLICOR results (two false-negative SRT-PCR results and one false-positive SRT-PCR result). The quantity of HCV RNA in sera from five (11%) patients with viremia detected by AMPLICOR was below the bDNA assay cutoff (< 3.5 x 10(5) RNA equivalents per ml). AMPLICOR compared favorably with SRT-PCR, with key advantages of speed, ease of use, increased sample throughput, and protection against false-positive results because of amplicon carryover.