The objective of the study was to compare the clinical sensitivity and specificity of versions 1.0 and 2.0 of the branched DNA (bDNA)-based hepatitis C virus (HCV) RNA quantification assay, and also to compare the values yielded by the two versions according to the HCV genotype. Serum samples from 268 patients tested routinely by a non-quantitative HCV RNA PCR assay (group A) were tested with version 2.0 of the bDNA assay. Samples from 342 HCV PCR-positive patients with chronic hepatitis C eligible for interferon treatment (group B) were tested with both version 1.0 and version 2.0 of the bDNA assay. Version 2.0 had a clinical sensitivity of 92% (95% confidence interval (CI): 87-97%) in group A and 89% (86-92%) in group B. In group B, the gain in sensitivity with bDNA 2.0 was 16% relative to bDNA 1.0 (P < 0.001). The log values of the two assays correlated with samples positive by both assays (r = 0.83, P < 0.0001), but the distribution of values was larger in samples containing HCV genotypes 2 and 3. The mean ratio of assay 2.0/assay 1.0 values was 1.69 +/- 1.44 (range: 0.33-13.43). The mean ratio was close to 1 with samples containing genotype 1 or 4, but ranged from 0.33 to more than 5. The mean ratio was close to 3 with samples containing genotype 2 or 3, and ranged from 0.5 to more than 13. HCV RNA levels were significantly lower in samples containing genotype 4 than in those containing other genotypes. Sera from 200 anti-HCV-negative, HCV RNA PCR-negative blood donors (group C), and from 164 anti-HCV-negative patients with symptoms of chronic liver disease (group D) were used to assess the clinical specificity of bDNA 2.0. In addition, samples with an HCV RNA titer between 0.2 (assay cutoff) and 0.5 MEq/ml from a group of 546 patients tested routinely for HCV RNA load by bDNA 2.0 (group E) were retested by bDNA 2.0 and by qualitative PCR. The specificity of bDNA 2.0 was 100% (98-100%) in group C and 99% (97-100%) in group D. Among the 41 samples from group E, 38 were positive by bDNA 2.0 retesting (36 were PCR-positive) and three were negative by bDNA 2.0 retesting (all were PCR-positive). It is concluded that version 2.0 of the bDNA assay is markedly more sensitive than version 1.0 and has a good specificity. In contrast with version 1.0, version 2.0 is not influenced by the HCV genotype. The relationship between values obtained with assays 1.0 and 2.0 on clinical specimens is not linear, indicating that HCV RNA titers cannot reliably be calculated from the results of version 1.0.