Objective:　To　explore　the　diversity　of　glycoside　hydrolase　family　10　xylanase　genes　in　the　sediment　of　soda　lake　Dabusu　by　using　high-throughput　amplicon　sequencing　based　on　the　Illumina　HiSeq2500　platform.　Results:　A　total　of　227,420　clean　reads,　representing　approximately　49.5 M bp,　were　obtained.　Operational　taxonomic　unit　(OTU)　classification,　with　a　95%　sequence　identity　cut-off,　resulted　in　467　OTUs　with　392　annotated　as　GH10　xylanase,　exhibiting　35–99%　protein　sequence　identity　with　their　closest-related　xylanases　in　GenBank.　Above　75%　of　the　total　OTUs　demonstrated　less　than　80%　identity　with　known　xylanases.　In　addition,　xylanases　derived　from　the　sediment　were　found　to　be　affiliated　to　12　different　phyla,　with　Bacteroidetes,　Proteobacteria,　Actinobacteria,　Firmicutes,　Verrucomicrobia,　and　Basidiomycota　being　the　dominant　phyla.　Moreover,　barcode　sequence　had　a　major　effect　on　abundance　with　only　a　minor　effect　on　diversity.　Conclusions:　High-throughput　amplicon　sequencing　offers　insight　into　xylanase　gene　diversity　at　a　substantially　higher　resolution　and　lesser　cost　than　library　cloning　and　Sanger　sequencing,　facilitating　a　more　thorough　understanding　of　xylanase　distribution　and　ecology.　©　2019,　Springer　Nature　B.V.