This perspective describes the concurrent development in the 1980s of the first transgenic mice genetically engineered to express dominant oncogenes, involving independent researchers who were largely unaware of each other's strategies and progress. We relate the experimental designs, the pitfalls and challenges encountered, and the eventual success in developing distinctive mouse models of cancer, wherein tumors arose heritably in various organs. These early oncomice have produced a wealth of new knowledge, become topics of intellectual property, and spawned a vibrant field of cancer research that is revealing mechanisms of tumorigenesis and suggesting new therapeutic strategies for treating the human disease.Supplemental material is available at http://www.genesdev.org.In the early 1980s, a technology for generating lines of "transgenic mice" carrying cloned genes integrated into the mouse genome was demonstrated to be a tractable and reproducible method (Gordon et el. 1980;Brinster et al. 1981;Costantini and Lacy 1981;Wagner et al. 1981; for review, see Palmiter and Brinster 1986). Concurrently, there was considerable excitement in cancer research, with the continuing discoveries and molecular cloning of viral and then cellular oncogenes. These genes were causally implicated in particular natural cancers and demonstrably capable of inducing transformation of cultured cells that would form tumors when transplanted in appropriate host animals. The two areas of research came together when transgenic mice carrying cloned oncogenes were generated and found in some cases to have heritable predispositions to the development of cancer. The first reports appeared in 1984, with others following in 1985-1987, collectively substantiating the hypothesis that oncogene expression in normal cells within normal tissues of a mammalian organism could lead to tumor development. In 1992, gene knockout technology converged in a similar fashion with tumor suppressor genetics in the generation of mice that developed cancers by virtue of lacking tumor suppressor gene function (for review, see Jacks 1996).The significance of these early technological innovations, which lead to genetically engineered mice endowed to heritably develop particular forms of cancer, can be appreciated by considering the scientific/technical landscape then, and now. Before these developments, cancer was largely modeled by tissue culture of cell lines established from human and animal tumors, and by the inoculation (transplantation) of such cell lines under the skin of immunodeficient mice, where lump-like solid tumors would form. While of clear utility in studying parameters of tumor growth, such models did not necessarily recapitulate the subtleties observed in human tumors arising in different organs, in terms of polymorphic genetic susceptibility, histological characteristics, and progression from benign premalignant lesions to tumors of increasing aggressiveness. Moreover, tumor transplant models, while frequently used as a benchmark to document activit...