Although HBV has the potential to generate an almost limitless spectrum of quasispecies during chronic infection, the viability of the majority of these quasispecies is almost certainly impaired due to constraints imposed by the remarkably compact organization of the HBV genome. On the other hand, single mutations may affect more than one gene and result in complex and unpredictable effects on viral phenotype. Better understanding of the constraints imposed by gene overlap and of genotype-phenotype relationships should help in the development of improved antiviral strategies and management approaches. Although the probability of developing viral resistance is directly proportional to the intensity of selection pressure and the diversity of quasispecies, potent inhibition of HBV replication should be able to prevent development of drug resistance because mutagenesis is replication dependent. If viral replication can be suppressed for a sufficient length of time, viral load should decline to a point where the continued production of quasispecies with the potential to resist new drug treatments no longer occurs. Clinical application of this concept will require optimization of combination therapies analogous to highly active antiretroviral therapy (HAART) for HIV infection. Total cure of hepatitis B will require elimination of the intranuclear pool of viral minichromosomes, which will probably only be achieved by normal cell turnover, reactivation of host immunity, or elucidation of the antiviral mechanisms operating during cytokine clearance in acute hepatitis B (see Fig. 1).