In　this　paper,　the　safety　threshold　of　asphalt　pavement　anti-sliding　performance　has　been　studied　based　on　drivers＇　psychological-physiological　reaction.　Time-series　theory　and　statistics　theory　were　used　to　analyze　drivers＇　electrodermal　activity　data.　The　results　showed　that　driver＇s　electrodermal　activity　(EDA)　data　was　detrended　linear　trend　first,　and　then　there　was　a　better　correlation　between　its　standard　deviation　(SD)　and　friction　coefficient　(1).　At　the　same　time,　there　was　also　a　better　relationship　between　SD　and　speed　when　driving　simulation　module　(DSM)　runs　at　a　specific　friction　coefficient　test　scene.　Based　on　these　conclusions　and　a　decay　curve　obtained　from　actual　road,　a　safety　threshold　of　f　was　acquired　and　the　optimal　safety　maintenance　opportunity　of　road　was　suggested.