We present evidence that C57BL/KsJ (BKs) arose through a genetic contamination of a black mouse strain. The majority of alleles in BKs are shared with C57BL/6J (B6). Regions differing from B6 share most alleles with DBA/2J (DBA). The allele distribution of typed markers indicates that the BKs genome is comprised of approximately 84% B6-1ike and 16% DBA-like alleles. The mouse strain C57BL/Ks has gained importance particularly in diabetes/obesity and atherosclerosis research. Hummel and coworkers described that the diabetes (db) mutant gene produces two distinct phenotypes depending on whether it is placed on a B6 or BKs background (Hummel et al. 1972). B6-db/db mice are hyperphagic, obese, hyperinsulinemic, but show only a mild diabetes with transitory hyperglycemia and hypertrophy of the islets of Langerhans. BKs-db/db mice share the hyperphagia, obesity, and hyperinsulinemia, but have a severe diabetes with marked hyperglycemia and beta-cell atrophy. A similar dichotomy is observed when the obese (ob) mutant gene is placed on both backgrounds (Coleman and Hummel 1973). Further studies indicated that the expression of the diabetic phenotype in the two inbred backgrounds is under multigenic control (Coleman and Hummel 1975). Nishina and associates showed that diet-induced formation of atherosclerotic lesions is more severe in a BKs than a B6 background (Nishina et al. 1994). For these reasons we were interested in the genetic origins of the BKs strain. The strain BKs can be traced back to a pair of reputed C57BL/6J mice that N. Kaliss (The Jackson Laboratory) obtained from JJ Biesele during a stay at the Sloan-Kettering Institute in New York. The male mouse was shipped from The Jackson Laboratory in October 1947, one day before the laboratory was destroyed in a forest fire. The female was one generation removed from the male and from a colony pen-bred by Biesele. In 1948, Kaliss brought the strain back with him to The Jackson Laboratory. During the course of successive inbreeding he tested the tissue rejection of a C57BL/6 tumor (E0771, sarcoma) in these mice. Initially the tumor grew in 100% of the mice; however, with continued inbreeding all mice finally rejected the tumor (N. Kaliss, letter to D. Dresser, NIMR, London, England, 1970). Because Kaliss observed only a black coat color in his BKs colony, at the time he favored the idea that a mutation occurred at the H2 locus, from H2 b (B6) to H2 a, as an explanation for the tumor rejection, rather than the idea that a genetic contamination had occurred. Over the following years more loci were found to be discordant between BKs and B6, rendering the mutation hypothesis less attractive. In 1982, one of us (DW Bailey) made an effort to resolve the issue. MATRIX (Roderick and Guidi 1989), a database containing the distribution of polymorphic loci among inbred mouse strains, was surveyed for the strain distribution patterns of inbred strains with the H-2 a haplotype at loci known to be discordant between BKs and B6. Most matches were found between BKs and NZB, fewer with BALB/cJ and DBA/2J. However, NZB cannot be the contaminating strain, because it had not yet been imported to the USA at the time when the contamination would have occurred,