The ability of mice to mount a cytotoxic T-lymphocyte (CTL) immune response to SV40 T-antigen is determined by the H-2 haplotype of the host; H-2b and k mice are high responders and H-2d mice are low responders. Mice of these 3 H-2 haplotypes were challenged with SV40 and their ability to generate and sustain an antibody response to SV40 T-antigen was found to be equivalent. To investigate the role of the different components of the host immune response in controlling growth of SV40-induced tumors, the tumorigenic potential of freshly established cell lines, obtained by SV40 transformation of cells from normal tissues of inbred strains of mice of 6 H-2 haplotypes, was assessed. Each cell line was tumorigenic in athymic and newborn mice but not in adult syngeneic immunocompetent mice. Cells from these initial SV40-transformed lines were then passaged in athymic (nu/nu) mice, re-established in vitro and again transferred into syngeneic animals. Transfer of H-2d SV40 transformants to low or non-responder mice of the H-2d haplotype resulted in tumor formation in some animals. Cells derived from these tumors expressed both the viral encoded T-antigen and the H-2Dd restriction element. Furthermore, the proportion of animals with tumors varied with the strength of their CTL-responsiveness to SV40 T-antigen in association with H-2Dd. Therefore, in H-2d animals, tumor cell growth appears to result from escape of cells from inefficient CTL surveillance. No tumors were formed by transfer of the in vivo selected H-2b or H-2k SV40 transformants to syngeneic high-responder mice. We therefore investigated the role of CTL in the selection of SV40-transformed cells able to escape immune surveillance. Under conditions of stringent immune selection by CTLs, tumorigenic cells that no longer expressed the relevant H-2 class-I restriction element were obtained. Although interaction between the various immune effector mechanisms may play a role in the recognition and elimination of SV40 transformants, our results were consistent with the hypothesis that the SV40-specific CTL response is the predominant control of SV40 tumor growth.