Abstract A territory of northern Russia covering more than 4,194 square kilometers is the world’s largest dwelling and breeding area of wild and domestic reindeer. Unlocking its complete genome sequences should have promoted design and development of species-specific genotyping arrays, but this process is still in progress. However, evaluation of state of gene pool of reindeer populations cannot be paused to prevent dramatic losses in genetic diversity. Thus, we aimed to select SNPs to create low-density panel to perform genetic assessments without losing biological content. Our study was based on SNP genotypes of wild reindeer (n = 83), and domestic reindeer from the Nenets District (Nenets breed, n = 100), Murmansk Region (Nenets breed, n = 19), Republic of Sakha (Yakutia) (Evenk breed, n = 19) obtained using Illumina BovineHD BeadChip. The data processing performed in PLINK 1.9 revealed 4456 polymorphic SNPs. Using Delta method implemented in the TRES software, 368 informative SNPs were selected from the whole set of polymorphic SNPs for further analysis. To compare the resolution power of low-density SNP panel with a whole set of polymorphic SNPs, we estimated minor allele frequencies (MAF) and performed PCA analysis and Admixture clustering. In case of low-density panel, we observed obvious bias to higher MAF (≥ 0.30), and 4% of SNPs had MAF around 0,1. Among the whole set of polymorphic SNPs, 70% SNPs had MAF ≤ 0.2. PCA obtained with low-density panel and the whole SNP set provided similar pattern of genetic differentiation between studied groups. Wild animals were clearly separated by PC1 from their domestic relatives. Besides Evenk reindeer was distant from two Nenets groups. The Admixture clustering demonstrated the identical patterns for both SNPs sets. The next step of our study will be developing custom DNA array based on 368 selected informative SNPs. The research was funded by RSF № 16-16-10068-P.