In short-term carcinogenicity testing using CB6F1-TgrasH2 mice, sibling nonTgrasH2 mice are used as a negative control. However, selection of TgrasH2 and nonTgrasH2 mice has been performed by PCR with only transgene specific primers by the conventional method. Therefore, the conventional method involves the risk of false negative results due to reaction failure, and contamination with TgrasH2 mice in the control mice group. Based on the nucleotide sequence information around the pre-integration site, we developed a genotyping method for distinguishing not only TgrasH2 mice (hemizygous for the Tg allele) but also nonTgrasH2 (homozygous for the nonTg allele) in a positive manner.