Oncogene-induced senescence is a cellular response that may protect against cancer development. However, its study has mainly focused on cultured cells with overexpressed oncogenes. This study analyzed tumors initiated by an endogenous oncogene, ras, and found that senescent cells exist in premalignant tumors but not in malignant ones. Senescence is thus a defining feature of premalignant tumors, potentially useful for cancer diagnosis and prognosis. A mouse model was used to study cancer initiation in humans. These mice have a conditional K-rasV12 allele activated by Cre recombinase, leading to multiple lung adenomas (premalignant) and a few adenocarcinomas (malignant). Senescence markers were used to detect oncogene-induced senescence in lung sections. Markers such as p16, p15, Dec1, and DcR2 were identified, as well as senescence-associated β-galactosidase and heterochromatin foci. Immunohistochemical analysis showed that adenomas were positive for these markers, while adenocarcinomas were negative. Proliferation markers showed a weaker index in adenomas compared to adenocarcinomas. Senescence-associated β-galactosidase and heterochromatin foci were also more prevalent in adenomas. These findings were confirmed using different Cre alleles and other methods. When the K-rasV12 allele was combined with a Cre allele targeting the pancreas, premalignant lesions were positive for senescence markers, while malignant ductal adenocarcinomas were not. Similarly, chemically induced skin papillomas were positive for these markers. The study suggests that oncogene-induced senescence may restrict tumor progression. In most cells, oncogenic K-ras signaling is not sufficient to trigger tumors or senescence. Rare cells that do not fully suppress oncogenic Ras may lead to both premalignant and malignant tumors. The study concludes that many cells in premalignant tumors undergo oncogene-induced senescence, but malignant tumors lack these effectors.Oncogene-induced senescence is a cellular response that may protect against cancer development. However, its study has mainly focused on cultured cells with overexpressed oncogenes. This study analyzed tumors initiated by an endogenous oncogene, ras, and found that senescent cells exist in premalignant tumors but not in malignant ones. Senescence is thus a defining feature of premalignant tumors, potentially useful for cancer diagnosis and prognosis. A mouse model was used to study cancer initiation in humans. These mice have a conditional K-rasV12 allele activated by Cre recombinase, leading to multiple lung adenomas (premalignant) and a few adenocarcinomas (malignant). Senescence markers were used to detect oncogene-induced senescence in lung sections. Markers such as p16, p15, Dec1, and DcR2 were identified, as well as senescence-associated β-galactosidase and heterochromatin foci. Immunohistochemical analysis showed that adenomas were positive for these markers, while adenocarcinomas were negative. Proliferation markers showed a weaker index in adenomas compared to adenocarcinomas. Senescence-associated β-galactosidase and heterochromatin foci were also more prevalent in adenomas. These findings were confirmed using different Cre alleles and other methods. When the K-rasV12 allele was combined with a Cre allele targeting the pancreas, premalignant lesions were positive for senescence markers, while malignant ductal adenocarcinomas were not. Similarly, chemically induced skin papillomas were positive for these markers. The study suggests that oncogene-induced senescence may restrict tumor progression. In most cells, oncogenic K-ras signaling is not sufficient to trigger tumors or senescence. Rare cells that do not fully suppress oncogenic Ras may lead to both premalignant and malignant tumors. The study concludes that many cells in premalignant tumors undergo oncogene-induced senescence, but malignant tumors lack these effectors.