In　this　paper　a　method,　based　on　the　cointegration　analysis　of　wind　turbine　SCADA　data,　is　proposed　to　solve　the　problems　such　as　misreporting　and　false　alarm　caused　by　affected　by　environmental　effect　in　wind　turbine　SCADA　system.　Using　cointegration　theory　to　the　wind　turbine　SCADA　data　will　produce　a　stationary　cointegration　residual　involving　state　information　of　wind　turbine.　Through　analysis　this　residual　can　realize　the　wind　turbine’s　condition　monitoring.　The　proposed　method　is　using　the　experimental　SCADA　data　as　the　research　object　what　is　acquired　from　a　1.5 MW　DFIG　under　varying　environmental　and　operational　conditions　in　Gu　Yang　County,　Bao　Tou　City,　Inner　Mongolia.　A　cointegration　model　was　established　using　some　of　the　nonstationary　data　and　the　model　was　validated　with　a　set　of　known　gearbox　fault　data.　The　result　demonstrates　that　the　proposed　method　can　effectively　restrain　the　response　caused　by　environment　and　operation　in　SCADA　data,　accurately　identify　the　running　state　of　wind　turbine,　and　simply　and　effectively　realize　the　state　monitoring　of　wind　turbine.　©　2020,　Springer　Nature　Switzerland　AG.