A modified competitive RT-PCR (mcRT-PCR) to measure HCV RNA in serum and the Amplicor HCV Monitor assay were compared. For mcRT-PCR, the RNA extracted was retrotranscribed and coamplified in one step with a known amount of a DNA internal control (IC). Digoxigenin-labeled amplified products were hybridized to specific HCV DNA and IC-DNA probes and quantified by colorimetry. HCV RNA concentration was calculated by plotting the ratio of HCV/IC ODs against a calibration curve. Multiple samples were analyzed in the same round and tedious titration of each sample with a competitor was unnecessary. The mcRT-PCR assay was linear from 6 x 10(3) to 6 x 10(7) copies/ml, whereas Amplicor was linear up to 1-2 x 10(6) copies/ml. HCV RNA was measured in samples from 75 carriers. There was agreement between both methods in type 1 infections but not in type 2 or type 3 infections, in which the values measured by Amplicor were, on average, 15 times lower than those measured by the mcRT-PCR. HCV RNA measured by Amplicor was higher in type 1 infections than in type 2 or 3 infections, but no differences were found when viral load was assessed by mcRT-PCR. The binding efficiency of the Amplicor-probe was greater for type 1 than for types 2 or 3, suggesting Amplicor underestimates the viral load in the latter types. In contrast, the mcRT-PCR is not affected by genotype-related variation of HCV. This study suggests that mcRT-PCR assay is reliable for sensitive and accurate measurement of HCV RNA over a broad range of values independently of the HCV genotype.