The ratio between wild-type hepatitis B virus (HBV) and HBV mutant, unable to secrete "e" antigen (HBeAg minus HBV) appears to be an important determinant of the outcome of chronic hepatitis B. Quantitative analysis of wild-type and HBeAg minus HBVs in the blood could be useful to monitor chronic hepatitis B patients. We developed a solid-phase minisequencing assay for both viruses using a primer-guided incorporation of a single labeled nucleotide on an affinity captured biotinylated amplified HBV-DNA template. A standard curve was constructed by mixing increasing quantities of wild type and mutant virus DNAs. The detection of wild-type and HBeAg minus sequences, ranging from 10% to 90% of overall viremia, was linear and reproducible till 0.1 pg/microliter of serum HBV-DNA. The assay yields numerical values and the ratio of incorporated nucleotides defines the relative proportions (%) of the two viral sequences with accuracy. We tested the sensitivity and accuracy of the minisequencing on mixed end point dilutions of wild-type and HBeAg minus reference sera and amplified products. The feasibility and reproducibility of the assay were tested in 35 sera from 21 HBsAg positive patients with chronic hepatitis B using both minisequencing and oligo-hybridization assays. A high correlation was found between the two assays (r = 0.957 P < 0.0001). In conclusion, the minisequencing assay provides a precise and reproducible quantitative analysis of wild-type and HBeAg minus HBVs in clinical specimens. It is proposed to study the relations between HBV heterogeneity and the course of hepatitis B and its response to therapy.