Soil　microorganisms　play　important　roles　in　the　ecosystem　functioning　of　subtropical　broad-leaved　forests　(SBFs).　However,　the　patterns　and　environmental　indicators　of　soil　microbial　functional　structure　remain　unclear　in　SBFs.　In　the　present　work,　we　used　a　functional　microarray　(GeoChip　4.0)　to　examine　the　soil　microbial　functional　structure　of　three　types　of　SBFs,　including　a　deciduous　broad-leaved　forest　(DBF),　a　mixed　evergreen-deciduous　broad-leaved　forest　(MBF),　and　an　evergreen　broad-leaved　forest　(EBF).　We　found　that　microbial　functional　structure　was　significantly　different　among　SBFs　(P　＜　0.05).　Compared　to　the　DBF　and　the　EBF,　the　MBF　had　higher　functional　a-diversity　(P=　0.001,　F=　12.55)　but　lower　beta-diversity　(P　＜　0.001,　F=　61.09),　and　showed　more　complex　functional　gene　networks.　Besides,　the　MBF　had　higher　relative　abundances　of　functional　genes　for　carbon　(C)　decomposition,　C　fixation,　nitrogen　(N)　cycling,　sulfur　(S)　cycling,　and　phosphorus　(P)　cycling　(P　＜　0.05),　indicating　stronger　microbial　functional　capabilities　of　nutrient　cycling　processes.　Edaphic　variables　(i.e.,　soil　pH　and　soil　nutrient　content)　were　revealed　as　better　indicators　of　soil　microbial　functional　structure　than　plant-related　ones　(i.e.,　vegetation　type　and　plant　diversity)　in　SBFs.　For　example,　functional　gene　structure　of　the　DBF　was　significantly　related　to　soil　total　S　(P　=　0.041),　that　of　the　MBF　was　significantly　related　to　soil　organic　C　(P　=　0.027)　and　plant　available　P　(P　=　0.034),　and　that　of　the　EBF　was　significantly　related　to　soil　pH　(P　=　0.006)　and　total　potassium　(K)　(P　=　0.038).　Overall,　through　the　analysis　of　microbial　functional　gene　profiles,　this　study　yields　unique　insights　into　the　environmental　indicators　of　patterns　and　mechanisms　of　soil　microbial　functional　structure　in　SBFs.　(C)　2021　Published　by　Elsevier　B.V.