The Himalayan marmot (Marmota himalayana) is an exceedingly common species widely-distributed across vast areas of the Qinghai-Tibet Plateau. In order to improve conservation and management strategies of this species, we must understand its genetic structure and genetic diversity at the population level. In this study, 258 Marmota himalayana individuals were collected from 13 geographic populations. The genetic diversity and genetic structures of these populations were analyzed according to the sequence variation of the mitochondrial D-loop sequence. All 258 samples were successfully sequenced for 887 bp of the mitochondrial D-loop sequence. We found 84 sites (22.49%) to be polymorphic and genetic diversity was high overall (mean haplotype diversity 0.968±0.003, mean nucleotide diversity 0.01725±0.01637). The results of AMOVA analysis showed obvious genetic differentiation (Fst=0.62067, P<0.001) and relatively low gene flow among different populations (Nm<1). A neighbor-joining phylogenetic tree produced two main branches; the first branch (A) included 18 haplotypes from the southwestern Qinghai-Tibet Plateau (including Tibet Andu, Qinghai Golmud, Qinghai Nangqian, and Yunnan Diqing), and the second branch (B) included the other 50 haplotypes from the northeastern Qinghai-Tibet Plateau. A similar topology was also found using network diagram methods. Populations of the Himalayan marmot on the Qinghai-Tibet Plateau are separable into two groups, with the Tanggula Mountains serving as a geographic barrier between them. Therefore, geographical isolation via this mountain range appears to be the primary factor enforcing genetic structure of Himalayan marmot populations.