The　operation　states　of　rolling　bearings　which　are　the　most　common　and　important　parts　in　the　mechanical　equipment,　will　affect　the　whole　machine　operation　condition　directly.　Due　to　the　working　environment　of　rolling,　bearing　is　complicated,　the　fault　vibration　signal　of　rolling　bearing　is　usually　non-stationary,　and　the　strong　noise　interference　is　contained　in　the　vibration　signal　at　the　same　time.　So　it　is　important　to　eliminate　the　noise　interference　and　extract　fault　feature　information　effectively　for　the　rolling　bearing.　Dual-tree　complex　wavelet　packet　transform　is　a　new　method　of　signal　processing.　Dual-tree　complex　wavelet　packet　transform　has　many　good　characteristics,　for　example,　approximate　shift　invariance,　good　directional　selectivity,　perfect　reconstruction,　limited　data　redundancy,　efficient　computational　efficiency　and　so　on.　The　high　frequency　part　of　dual-tree　complex　wavelet　transform　that　is　not　decomposed,　is　further　decomposed　by　dual-tree　complex　wavelet　packet　transform,　so　as　to　improve　the　whole　frequency　band　signal　frequency　resolution　and　reduce　the　loss　of　information.　In　view　of　the　above　situation,　a　new　fault　diagnosis　method　is　proposed　based　on　dual-tree　complex　wavelet　packet　transform　and　threshold　de-noising.　Firstly,　the　non-stationary　fault　signal　is　decomposed　into　several　different　frequency　band　components　through　dual-tree　complex　wavelet　packet　decomposition.　Secondly,　Kurtosis　and　the　cross-correlation　coefficient　of　each　component　are　obtained　and　compared.　Due　to　the　kurtosis　reflecting　the　signal　variations,　if　the　kurtosis　value　is　bigger,　the　degree　of　the　change　of　signal　is　bigger　too.　The　correlation　coefficient　can　reflect　the　proximity　between　the　component　and　the　original　signal　at　the　same　time,　the　correlation　coefficient　is　bigger,　the　more　similar　with　the　original　signal.　Finally,　the　components　that　have　a　bigger　value　are　chosen　to　be　de-noised　by　a　soft　threshold　and　reconstructed　by　dual-tree　complex　wavelet　packet　transform.　The　noise　interference　was　eliminated　effectively,　and　the　effective　signal　information　was　retained　at　the　same　time.　Thus,　the　fault　feature　information　was　extracted.　In　this　paper,　the　outer　bearing　fault　test　and　the　engineering　practical　fault　data　of　rolling　bearing　were　analyzed　by　dual-tree　complex　wavelet　packet　transform　and　threshold　de-noising.　In　order　to　contrast　analysis,　the　vibration　signals　were　processed　by　the　traditional　discrete　wavelet　packet　transform　and　threshold　de-noising.　The　signal　has　the　better　periodic　impact　obtained　by　the　method　proposed　in　this　paper,　and　the　noise　was　eliminated　ideally.　So　the　fault　diagnosis　method　based　on　dual-tree　complex　wavelet　packet　transform　and　threshold　de-noising　can　effectively　eliminate　the　noise　in　the　vibration　signals.　Thus,　the　fault　feature　information　was　extracted　and　the　feasibility　and　effectiveness　of　this　method　is　verified.