To investigate the possibility of using hepatitis B virus (HBV) as a vector, the tat gene from human immunodeficiency virus type 1 (HIV-1) was inserted into the full-length HBV genome in-frame with the polymerase (pol) open reading frame in the tether region and downstream of the preS1 promoter. We demonstrated that the tat gene was expressed with full activity in transactivating the HIV-1 long terminal repeat (LTR). The expression of the tat gene in the context of the HBV genome in chicken hepatoma and human cervical carcinoma cells, however, was not as efficient as that in human hepatoblastoma cells, which reflects the cellular and species specificity of promoters of hepadnaviruses. Detection of RNA expressed from this HBVtat recombinant revealed transcription of the tat gene by two promoters: the core/pol promoter and the preS1 promoter. A Pol-Tat fusion protein expressed by the core/pol promoter did not seem to contribute to the tat transactivation activity of the HBVtat recombinant since a frameshift mutation in the pol gene did not affect the recombinant tat function. The functional tat protein, therefore, was most likely expressed as a Tat-Pol fusion product. Endogenous polymerase assays showed that the pol protein expressed from the HBVtat recombinant was still active although at a reduced level. Hepatitis B surface antigens and e antigen produced from this recombinant were detected at similar levels as those produced from the wild type. Notably, the capability of forming complete HBV particles was still retained. These studies indicate the potential of constructing HBV as a replicative vector. We also showed that manipulation of a nonreplicative HBV vector was possible. Expression of the HBV polymerase could be completely eliminated and replication of the nonreplicative HBV recombinant could be supported by Pol transcomplementation.