In　high　efficiency　video　coding　(HEVC),　up　to　35　intra　prediction　modes　are　adopted　to　improve　encoding　performance.　However,　large　numbers　of　modes　lead　to　increased　coding　complexity　inevitably,　which　hinders　the　promotion　of　HEVC　in　real-time　environment.　To　reduce　the　encoding　complexity,　a　spatial-temporal　correlation　based　fast　intra　mode　decision　algorithm　is　proposed　in　this　paper.　Herein,　the　rough　mode　decision　(RMD)　candidate　list　and　the　rate　distortion　optimization　(RDO)　candidate　list　are　narrowed　respectively　by　strong　spatial　correlation　and　temporal　correlation　between　prediction　units　(PUs).　For　RMD,　the　proposed　algorithm　early　determines　the　impossible　range　of　the　PU　optimal　mode　in　line　with　spatial　correlation,　which　would　narrow　RMD　candidate　list　and　speed　up　RMD　process.　For　RDO,　the　proposed　algorithm　tries　to　combine　temporal　correlation　with　intra　coding　to　simplify　RDO　candidate　list　and　save　RDO　time　consuming.　Experimental　results　demonstrate　that　the　proposed　algorithm　yields　average　28.64%　encoding　time　reduction　with　1.19%　BDBR　gain　or　0.06dB　BDPSNR　loss　compared　with　HM16.9.　The　proposed　algorithm　drives　HEVC　coding　efficiency　under　real-time　environments.　©　2018,　Ubiquitous　International.　All　rights　reserved.