The risks of secondary lung cancer in patients with early stage non-small and small cell lung cancers are estimated to be 1-2% and 2-10% per patient per year, respectively. Surprisingly, the incidence of second primary cancer in locally advanced non-small cell lung cancer at 10 years, following cisplatin-based chemotherapy with concurrent radiotherapy, increases to 61%. Those patients, on the road to being cured, cannot overlook the possibility of developing a second primary cancer. We developed a second primary lung cancer model using cisplatin as a carcinogen in A/J mice to screen for chemopreventive agents for a second malignancy. In the primary lung tumour model, 4-(methyl-nitrosamino)-1-(3-pyridyl)-1-butanone (NNK), benzo(a)pyrene (BaP), urethane induces specific K-ras mutations in codon 12, codon 12, and codon 61, respectively, in the A/J mice. In this study, we investigated the mechanisms of carcinogenicity by cisplatin in the A/J mice. In the cisplatin-induced tumours, we found no K-ras codon 12 mutation, which is the major mutation induced by NNK or BaP. K-ras gene mutations in codon 13 and codon 61 were found in one tumour (4%) and five tumours (17.8%), respectively. These findings suggest that cisplatin is partially related to K-ras codon 61 mutations, and that the mechanism of carcinogenicity by cisplatin is different from that by NNK or BaP.