To compare hepatitis C virus (HCV) RNA levels determined by branched DNA probe assay and by competitive polymerase chain reaction (PCR) as predictive markers of the response to interferon for treatment of patients with chronic HCV infection, we studied data on 140 patients treated for six months with natural interferon-alpha. Serum samples were tested for HCV RNA by PCR. HCV RNA was grouped into four genotypes by PCR with type-specific primers, and HCV RNA level was measured by branched DNA probe assay and by competitive PCR. HCV RNA was detected in all patients prior to initiation of the treatment. A complete response, sustained elimination of HCV RNA, occurred in 51 patients (36.4%). With multiple logistic regression analysis assessment, when using competitive PCR, a low level of HCV RNA (P < 0.0001), younger age (P = 0.0054) and genotype 2a and 2b (P < 0.0158) were significant predictive markers for a complete response to interferon treatment. When using branched DNA probe assay, a low level of HCV RNA (P < 0.0001) and age (P = 0.0089) were predictive markers, but genotype was not. The branched DNA probe assay had a narrower linear range for quantitation of HCV RNA level than competitive PCR. In conclusion, HCV RNA level determined by branched DNA probe assay proved to be useful for prediction of effects of interferon and it is cost effective as a marker of complete response to interferon treatment for patients with chronic HCV infection.