Ruhlen's hypothesis, based on linguistic evidence, for a common phylogenetic origin of Na-Dene and Yeniseian speakers is tested using genetic data. Gene frequency data for the Kets, the only surviving Yeniseian speakers, were collected during a field study in 1993. Data for several Na-Dene groups, as well as other New World and Siberian populations, were compiled from the literature. These data were analyzed using R-matrix, principal components analysis, and Mantel tests. In a comparison of 10 New World and Siberian populations using eight alleles, 55.8% of the variation was accounted for by the first principal component, and 22.1% of the variation was subsumed by the second principal component. Contrary to Ruhlen's interpretation of the linguistic data, analysis of the genetic data shows that the Na-Dene cluster with other Native American populations, while the Kets genetically resemble the surrounding Siberian groups. This conclusion is further supported by correlations that are higher when the Kets are considered unrelated to Na-Dene speakers, and an insignificant partial correlation between genes and language when geography is held constant, indicating that spatial patterning accounts for most of the variation present in these populations.