All of us are involved in discovery science as we pursue the genes, networks, cellular processes and biophysical principles that govern our chosen biological question. For those of us who choose to proceed using plant or animal models to dissect the elements of our favorite biological system, there are many classical and newer approaches available for our use, including two complementary strategies by which the discovery process is proceeding at the Oak Ridge National Laboratory (ORNL). The ORNL has been known for six decades for its investigations of the effects of radiation and chemicals in inducing heritable mutations in mouse germ cells, and for using mouse mutations as tools for the cloning and characterization of mammalian genes. Our history and experience in making mouse models are being applied via these two complementary strategies: 1), a phenotype-driven approach, in which mice carrying random chemically-induced mutations are screened for abnormal phenotypes; and 2) a gene-driven approach in which heritable single nucleotide polymorphisms (SNP) in preselected genes already thought likely to influence a biological system of choice can be recovered in live mice. The SNP-carrying mice can then be phenotyped for alterations in one's target biology. Both approaches have value and are necessary; while we can use mutations in genes that we already know to be of interest in our favorite biology to discover gene function, we also know that biology is full of surprise genes whose effects on our favorite biology would not be predicted and which will be identified only through phenotype screening.