The polymerase chain reaction (PCR) is used to detect hepatitis C virus (HCV) RNA, and the results of this assay may have a bearing on management of patients. We tested 31 laboratories for performance of HCV PCR with a coded panel that comprised 4 HCV-positive plasma samples, 6 HCV-negative samples, and two dilution series of HCV-positive plasma. 15 (48%) laboratories had faultless results with both dilution series, and 16 (52%) laboratories reported erroneous results with one or both series. 10 (32%) laboratories had faultless results when testing undiluted plasma samples, 11 (35%) produced a false-negative result with a weak-positive sample, and 10 (32%) produced false negative and/or false positive results. Only 5 (16%) laboratories performed faultlessly with the entire panel of samples. Reports of presence of HCV should be interpreted with care until reliable HCV-RNA detection becomes widely available.