Since new guidelines on ICH carcinogenicity testing of pharmaceuticals have been adopted, evaluation on carcinogenicity of newly-developed pharmaceuticals using genetically modified animals has been conducted. Validation studies have suggested that rasH2 mice carrying human prototype c-Ha-ras gene, heterozygous p53 deficient [p53 (+/-)] mice and homozygous XPA deficient [XPA (-/-)] mice are very susceptible to genotoxic carcinogens. In addition, many experimental studies using rasH2 and p53 (+/-) mice in our laboratory suggest that these mouse models are very useful for clarifying the mechanism of organ-specific carcinogenesis. On the other hands, there are unsolved mechanisms and points that need to be clarified in these models, as described below. (1) The mechanism of the enhanced carcinogenesis is not completely understood; point mutations of human prototype c-Ha-ras gene are not always related to the tumor induction in rasH2 mice, but recent studies demonstrated that overexpression of the transgene is responsible for the enhanced carcinogenesis. (2) Malignant lymphomas are induced by the treatment of phenolphthalein in p53 (+/-) C57Bl/6 mice, in which exon 5 of the lateral p53 allele was inactivated, but not in p53 (+/-) CBA and C57Bl/6 mice, in which exon 2 of the p53 allele was inactivated. In this respect, such a strain difference and different genetically engineering procedure should be taken into account when p53 (+/-) mice are planned to be used for the evaluation of carcinogenic potential of newly-developed chemicals.