This study examines the craniometric differentiation of Northern Eurasian groups with reference to genetic and partly linguistic facts. Measurements of 66 series of male crania from that territory, dating to various periods from the Mesolithic to the Early Bronze Age, were subjected to statistical methods especially destined for detecting spatial patterns, specifi cally gradients. Using the nonmetric multidimensional scaling of the matrix of D2 distances corrected for sample size, a two-dimensional projection of group constellation was generated, and a minimum spanning tree, showing the shortest path between group centroids in the multivariate space, was constructed. East-west clines in Northern Eurasia, detected by both genetic and craniometric traits, likely indicate not so much gene fl ow as isolation by distance, resulting from an incomplete evolutionary divergence of various fi lial groups constituting the Boreal meta-population. The western fi lial component, which, in Siberia and Eastern Central Asia, is mostly represented by Afanasyevans, has evidently made little contribution to the genetic makeup of later populations. The eastern fi lial component, which had appeared in the Cis-Baikal region from across Lake Baikal no later than the Neolithic, admixed with the autochthonous Paleosiberian component. The latter’s principal marker—the ANE autosomal component—had been present in Siberia since the Upper Paleolithic. Likewise autochthonous were both Eurasian formations—Northern and Southern; statis tical analysis has made it possible to make these more inclusive, whereby the former has been expanded in the eastern direction to include the Kuznetsk Basin, and the latter westwards, to the Middle Irtysh. Nothing suggests that Eastern European groups had taken part in the origin of either the Northern Eurasian formation or the proto-Uralic groups.