Tenofovir, an antihuman immunodeficiency virus (HIV) drug, has activity against lamivudine-resistant hepatitis B virus (HBV) mutants. To describe the efficacy of tenofovir in patients with lamivudine-resistant hepatitis B we applied two investigative approaches based on mathematical models of viral dynamics: the individual nonlinear fitting and the mixed-effect group fitting approaches. Eleven chronic HBV patients on lamivudine for a median of 176 weeks (range: 72-382) with YMDD mutation-related HBV-DNA breakthrough received 'add-on' tenofovir 300 mg once-daily, while maintaining their existing therapy. Sequential sera were taken at day 1 (t = 0 and t = 8 h), days 2, 4, 7, 10, 14, 21, 28 and every 4 weeks thereafter, and HBV-DNA levels were assessed using a validated quantitative polymerase chain reaction (PCR) assay. Median baseline log HBV-DNA was 8.62 (range: 6.48-9.76 log HBV-DNA). Tenofovir treatment resulted in a mean (+/-SD) log HBV-DNA decline of 1.37 +/- 0.51 in the first phase, 2.54 +/- 0.91 after 4 weeks, and 4.95 +/- 0.90 log HBV-DNA after 24 weeks. The median effectiveness of blocking viral replication in the individual fit model was 93% (range: 73-99) for eta = 0 and 93% (range: 59-99) for eta = 1. There was only a small difference between the efficacy parameter 'epsilon' of the individual nonlinear fitting and mixed-effect group fitting on the biphasic exponential model. These data show that tenofovir has good efficacy in blocking viral replication in HBV patients with lamivudine-induced drug-resistant HBV mutants, but effectiveness varies greatly among individuals. Both models can be used to describe viral decay during tenofovir therapy.