Tool　wear　is　one　of　the　most　common　tool　failure　modes.　And　the　vibration　signals　of　tool　wear　show　strong　nonlinear　and　non-Gaussian　characteristics.　Bispectrum　is　an　effective　method　to　analyze　nonlinear　and　non-Gaussian　signals.　However,　bispectrum　is　not　only　difficult　to　understand,　but　also　computationally　intensive.　This　paper　presents　a　milling　tool　condition　monitoring　method　based　on　the　amplitude　feature　of　tool　characteristic　frequency　of　bispectral　diagonal　slices.　Bispectral　diagonal　slice　is　used　as　tool　wear　signal　processing　method　to　identify　tool　states.　And　the　amplitude　feature　of　tool　characteristic　frequency　of　bispectral　diagonal　slices　was　selected　to　monitor　tool　wear.　The　effectiveness　of　the　proposed　method　has　been　verified　by　the　experiment.　Results　showed　that　the　proposed　approach　is　able　to　effectively　extract　the　feature　information　of　tool　wear　state,　and　reduces　the　calculation　cost.　©　2021　IEEE.