Two tests for measurements of hepatitis B virus DNA in human serum and plasma, VERSANT HBV 3.0 (based on the use of branched DNA chains) and Biotitre-B (a real-time PCR variant), were compared. Serum specimens from patients (n=56) with documented viral hepatitis were tested. For specimens with DNA values in the linear range in both tests the correlations were evaluated by Pearson's method. The sensitivity of the two tests, and reproducibility of the results of HBV DNA measurements were evaluated using a panel of recombinant virus DNA dilutions with a step of 1 lg HBV DNA copies per ml (from 7 to 3 lg). Three measurements for each concentration of the reference sample were carried out in each test. Clinical specificity of the two tests was evaluated by the analysis of HBV-negative serum samples, collected from donors not reactive by HBsAg and anti-HBc (n=60). Of 56 samples with documented infection, the results of testing by VERSANT HBV 3.0 and Biotitre-B did not agree in 4 (7.1%) samples. Pearson's correlation coefficient of for results obtained in VERSANT HBV 3.0 and Biotitre-B in linear range for both tests was 0.712. Evaluation of reproducibility of the tests using a panel of recombinant HBV DNA showed higher reproducibility of VERSANT HBV 3.0 test with coefficient of variations from 0.79 to 2.79% vs. 2.39-10.69% for Biotitre-B reference test. All 60 serum samples from donors areactive by HBsAg and anti-HBc were negative by HBV DNA when tested by VERSANT HBV 3.0 and Biotitre-B. Hence, clinical specificity of both tests was 100%. The results indicate high specificity of both tests and good agreement of their results, the reproducibility of VERSANT HBV 3.0 test being higher.