Most phylogenetic systematists assume speciation results in dichotomously branching phylogenies. Hybridization that gives rise to a new Lineage can produce character homoplasy that might obscure a species' true history. We report the results of a restriction-enzyme analysis of mitochondrial DNA (mtDNA) variation in three tiger salamander subspecies (Ambystoma tigrinum mavortium, Ambystoma tigrinum nebulosum, and Ambystoma tigrinum stebbinsi) and compare the results to studies of morphological and allozymic variation in these taxa. Allozymically, A. t. mavortium and A. t. nebulosum share most of their genomes (although each has several unique alleles), yet color pattern and mtDNA haplotypes are distinct. Color pattern and allozyme data suggest that A. t. stebbinsi shares a common ancestor with A. t. mavortium, while the A. t. stebbinsi mtDNA haplotype is derived from an A. t. nebulosum haplotype. Thus, our data suggest that A. t. stebbinsi originated through hybridization between A. t. mavortium and A. t. nebulosum. That hybridization can produce recognizably distinct evolutionary entities has long been recognized for plants, but the evolutionary significance of hybridization in animals should be examined more closely. Conservation agencies must recognize that hybrids and hybrid tars are not necessarily evolutionary 'mistakes,' and they might have significant importance in the production of natural biodiversity.