Background: Chaka sheep, named after Chaka Salt Lake, are adapted to a harsh, highly saline environment. They are known for their high-grade meat quality and are a valuable genetic resource in China. Furthermore, the Chaka sheep breed has been designated a geographical symbol of agricultural products by the Chinese Ministry of Agriculture. Results: The genomes of 10 Chaka sheep were sequenced using next-generation sequencing, and compared to that of additional Chinese sheep breeds (Mongolian: Bayinbuluke and Tan; Tibetan: Oula sheep) to explore its population structure, genetic diversity and positive selection signatures. Principle component analysis and a neighbor-joining tree indicated that Chaka sheep significantly diverged from Bayinbuluke, Tan, and Oula sheep. Moreover, they were found to have descended from unique ancestors (K = 2 and K = 3) according to the structure analysis. The Chaka sheep genome demonstrated comparable genetic diversity from the other three breeds, as indicated by observed heterozygosity (Ho), expected heterozygosity (He), runs of homozygosity (ROH), linkage disequilibrium (LD) decay. The enrichment analysis revealed that in contrast to Mongolian or Tibetan lineage groups, the genes annotated by specific missense mutations of Chaka sheep were enriched with muscle structure development (GO:0061061) factors including insulin-like growth factor 1 (IGF1), growth differentiation factor 3 (GDF3), histone deacetylase 9 (HDAC9), transforming growth factor beta receptor 2 (TGFBR2), and calpain 3 (CAPN3), among others. A genome-wide scan using Fst and XP-CLR revealed a list of muscle-related genes, including neurofibromin 1 (NF1) and myomesin 1 (MYOM1), under potential selection in Chaka sheep compared with other breeds. Conclusions: The comprehensive genome-wide characterization provided the fundamental footprints for breeding and management of the Chaka sheep and confirmed that they harbor unique genetic resources.