To　extract　the　weak　fault　information　submerged　in　strong　background　noise　of　the　bearing　vibration　signal,　multiwavelet　denoising　method　with　adaptive　threshold　and　envelope　demodulation　method　are　applied　in　this　paper.　Due　to　several　scaling　functions　and　wavelet　functions,　multiwavelets　have　many　excellent　properties　that　single　wavelet　cannot　satisfy　simultaneously,　such　as　symmetry,　orthogonality,　compact　support,　and　high　vanishing　moments,　which　make　it　match　different　characteristics　of　analyzed　signal.　GHM　multiwavelet　and　Db2　wavelet　are　used　to　analyze　the　simulated　outer　race　fault　signal　of　rolling　bearings,　in　which　adaptive　threshold　selection　strategy　is　introduced　in　multiwavelet　denoising.　Based　on　the　comparison　of　denoising　effects,　multiwavelet　adaptive　threshold　denoising　is　much　more　effective　than　single　wavelet.　Furthermore,　multiwavelet　denoising　method　is　applied　to　experimental　signal　and　engineering　data　individually.　Results　show　that　the　denoising　method　can　identify　the　incipient　fault　feature　as　early　as　possible,　which　cannot　be　realized　by　single　wavelet.