summary Prediction of treatment response is clinically important in chronic hepatitis C virus (HCV) genotype 4 infection. Early viral kinetics is useful in this respect for genotype 1 but interpretation is dependent on assay linearity and reproducibility. The VERSANT HCV RNA 3.0 (bDNA-3.0) and the COBAS Amplicor HCV Monitor 2.0 (HCM-2.0) have been widely used quantitative assays. We wanted to comparatively evaluate the two tests in a large genotype 4 sample. Genotyping was performed by NS5b sequencing. Viral load was tested in parallel in 32 patients at least six times on antiviral therapy with interferon α (IFNα). Totally, 198 samples within a quantitative range from undetectable to about 7 × 106 IU/mL (bDNA-3.0) were obtained and compared. Twenty-two samples with viral load above 500 000 IU/mL tested by HCM-2.0 were 1:100 diluted and retested. Quantitative values were fitted to a third order polynomial (M = 0.118303 + 1.07503 × V+ 0.0112128 × V2 − 0.0055504 × V3; M…HCM-2.0, V…bDNA-3.0, both log IU/mL) showing progressive nonlinearity of HCM-2.0 above 100 000 IU/mL but better clinical sensitivity with respect to bDNA-3.0. Dilution lead to a gain of at least a factor of 2.7 and thus, overestimation compared with bDNA-3.0. Deviation from linearity and overestimation upon dilution by HCM-2.0 are similar with HCV genotype 4, compared with other HCV genotypes. Differences in test performance were not detected for subtypes but for individual patients possibly related to specific quasispecies patterns. The interpretation of viral kinetic data becomes difficult due to overestimation upon dilution of baseline values by HCM-2.0.