The DNA repair-deficient Xpa(-/-)p53(+/-) (Xpa/p53) mouse is a potent model for carcinogenicity testing, representing increased sensitivity toward genotoxic but surprisingly also toward true human non-genotoxic carcinogens. The mechanism of this increased sensitivity in Xpa/p53 mice toward non-genotoxic carcinogens is still unknown. Here, we investigated the mechanism of the human non-genotoxic carcinogen cyclosporine A (CsA) in the Xpa/p53 mouse model. Xpa/p53 mice exposed to CsA for 39 weeks showed a significantly increased lymphoma incidence as compared with untreated Xpa/p53 mice and CsA-treated wild-type (WT) mice. We excluded concealed genotoxicity of CsA in Xpa/p53 mice by mutant frequency analyses. As a next step, we used a genetic approach: immunodeficient DNA-PKcs mice, defective in the catalytic subunit of the DNA-dependent protein kinase, were crossed with Xpa and Xpa/p53 mice. Xpa/p53 mice had an increased lymphoma incidence with shorter latency times as compared with DNA-PKcs-deficient WT and Xpa mice. Surprisingly, also six of 15 DNA-PKcs/Xpa/p53 females had developed an adenocarcinoma of the mammary gland. Tumor responses in CsA-treated and DNA-PKcs-deficient Xpa/p53 mice were comparable as both genotypes developed mainly splenic lymphomas enriched in B lymphocytes. From our present studies, we hypothesize that levels of initiated precancerous cells are elevated in Xpa/p53 mice. These cells are insufficiently eliminated due to either suppression of the immune system by CsA or through immune-related DNA-PKcs deficiency. Based on the current studies and those conducted previously, we conclude that the Xpa/p53 model is an excellent adjunct to the current chronic rodent bioassay.