Nearly all the experimental mice used in aging research are derived from lineages that have been selected for many generations for adaptation to laboratory breeding conditions and are subsequently inbred. To see if inbreeding and laboratory adaptation might have altered the frequencies of genes that influence life span, we have developed three lines of mice (Idaho [Id], Pohnpei [Po], and Majuro [Ma]) from wild-trapped progenitors, and have compared them with a genetically heterogeneous mouse stock (DC) representative of the laboratory-adapted gene pool. Mean life span of the Id stock exceeded that of the DC stock by 24% (P < 0.00002), and maximal life span, estimated as mean longevity of the longest-lived 10% of the mice, was also increased by 16% (P < 0.003). Mice of the Ma stock also had a significantly longer maximal longevity than DC mice (9%, P = 0.04). The longest-lived Id mouse died at the age of 1450 days, which appears to exceed the previous longevity record for fully fed, non-mutant mice. The life table of the Po mice resembled that of the DC controls. Ma and Id mice differ from DC mice in several respects: both are shorter and lighter, and females of both stocks, particularly Id, are much slower to reach sexual maturity. As young adults, Id mice have lower levels of insulin-like growth factor 1 (IGF-I), leptin, and glycosylated hemoglobin compared with DC controls, implicating several biochemical pathways as potential longevity mediators. The results support the idea that inadvertent selection for rapid maturation and large body size during the adaptation of the common stocks of laboratory mice may have forced the loss of natural alleles that retard the aging process. Genes present in the Id and Ma stocks may be valuable tools for the analysis of the physiology and biochemistry of aging in mice.