Resistance genes coding for inhibitors of hepadnaviral replication, such as ribozymes, antisense RNA, and dominant negative mutants have been shown to be effective in transfected hepatoma cells. In vivo studies, however, are not available to date. Here we expanded the use of the duck hepatitis B virus (DHBV) model for studying antiviral resistance genes in vivo. Animals were experimentally infected by intravenous injection of DHBV-positive serum in ovo. The use of recombinant human adenovirus type 5 and avian adenovirus CELO for gene transfer was evaluated. Adenovirus type 5 transduced more than 95% and CELO less than 1% of embryonic hepatocytes in vivo. Adenovirus type 5 interfered with DHBV replication (viral cross-talk), but this effect was moderate and did not preclude analysis of specific antiviral effects. Thus adenoviral transfer of a dominant negative mutant prior to DHBV infection (intracellular immunization) yielded 100-fold suppression of viral replication compared to the green fluorescent protein marker gene. Neither gene was toxic. These data demonstrate that a prototype anthepadnaviral resistance gene is functional in vivo. Duck embryos represent a useful model for evaluating gene therapeutic strategies in vivo without the need for large scale preparations of gene delivery vehicles.