The　rapid　development　of　China’s　logistics　industry　is　accompanied　by　the　deterioration　of　the　ecological　environment　and　excessive　energy　consumption.　Therefore,　how　to　effectively　measure　and　improve　the　green　total　factor　productivity　(GTFP)　of　the　logistics　industry　is　an　important　guarantee　for　achieving　the　coordination　of　the　logistics　industry　development　and　the　ecological　environment　protection　in　the　high-quality　development　stage.　This　study　evaluated　the　logistics　industry’s　GTFP　of　30　provinces　in　China　from　2004　to　2017　using　the　Epsilon-based　measure　model　(EBM)　and　global　Malmquist-Luenberger　index　(GML).　Then,　this　paper　applied　the　geographically　and　temporally　weighted　regression　(GTWR)　to　analyze　the　spatiotemporal　non-stationarity　of　influences　of　driving　factors　on　GTFP.　There　are　three　main　conclusions　drawn　in　this　paper.　Firstly,　the　GTFP　of　the　logistics　industry　has　significant　spatial　and　temporal　differences.　From　a　temporal　perspective,　the　GTFP　has　undergone　a　process　of　alternating　changes　in　ascent　and　descent.　From　a　spatial　perspective,　the　GTFP　has　an　obvious　“east-central-west”　gradient　decreasing　trend.　Secondly,　compared　with　the　ordinary　least　squares　(OLS)　and　the　geographically　weighted　regression　(GWR),　GTWR　performs　best　in　terms　of　goodness　of　fit.　Thirdly,　the　regression　results　of　GTWR　indicate　that　the　influences　of　factors　have　different　directions　and　intensities　on　GTFP　in　the　logistics　industry　at　different　times　and　regions,　showing　obvious　characteristics　of　spatiotemporal　non-stationarity.　Finally,　some　practical　recommendations　are　put　forward　in　this　paper.　©　2020,　HARD　Publishing　Company.　All　rights　reserved.