The correlation between 3 assays for hepatitis C virus (HCV) RNA quantification and their respective accuracy in predicting the response to interferon and interferon/ribavirin therapy was evaluated by analysing pre-treatment sera from 100 patients. A total of 97%, 100%, and 98% of the patients tested positive by the branched DNA 2.0 assay (Quantiplex), a multi-cycle reversed transcriptase polymerase chain reaction quantitative assay (Superquant) and the Roche Amplicor Monitor assay, respectively. The correlations between the assays, in all patients and in the major genotypes 1, 2, and 3, were significant, although the levels detected by the Amplicor Monitor assay were more than 1 log lower than by the other assays. Sustained virological responders to interferon therapy, but not to combination therapy, had lower baseline viral levels than long-term non-responders (p = 0.002 by Quantiplex 2.0; p = 0.008 by Superquant; p = 0.06 by Roche Amplicor Monitor). Pre-treatment viral load greater than 3 x 10(6) Eq or copies/ml by the Quantiplex 2.0 and Superquant assays and greater than 100,000 copies/ml by the Amplicor Monitor assay predicted long-term non-response in 94%, 93% and 91% of the interferon treated patients, respectively. In conclusion, acceptable correlations between available commercial quantitative assays were found. High baseline viral load predicted long-term non-response to interferon monotherapy, whereas it did not to interferon/ribavirin combination therapy.