Previous work has suggested that morphological transformation of cultured cells by human adenoviruses of subgroups A, B, and C is predominantly a function of early region 1b (E1b), and that region E1a has a role in immortalization. To test the hypothesis that region E1b is essentially responsible for the induction of the transformed phenotype the transforming activity of region E1b in the absence of region E1a was reinvestigated. In agreement with previous results, region E1b had no detectable transforming activity in primary baby rat kidney (BRK) cells nor in established rat cell lines. Since recent experimental evidence indicates that expression of E1b is blocked by a cellular factor which is inactivated by region E1a products, the regulatory signals in front of the coding sequence of region E1b were removed and replaced by the early promoter of SV40. These E1b-SV40pr plasmids had no detectable transforming activity in primary BRK cells, but they transformed normally in the presence of region E1a plasmids, demonstrating that both subregions are required for complete transformation of primary BRK cells. Transfection of the established rat cell line 3Y1 with the E1b-SV40pr plasmids did not result in complete morphological transformation either. Cotransfection of 3Y1 cells with E1b-SV40pr plasmids and pAG60 (a plasmid which harbors the kanamycin-resistance gene of Tn5) resulted in the appearance of foci of cells resistant to the antibiotic G-418. These colonies expressed the region E1b polypeptides to levels comparable to those found in cells transformed with intact region E1. Despite the presence of the E1b proteins the cells appeared essentially untransformed, in contrast to foci obtained after cotransfection of 3Y1 cells with mixtures of p5XhoI C (comprising region E1 DNA) and pAG60. These results indicate that complete transformation is a function of both regions E1a and E1b and that region E1a must have an important role in morphological transformation.