As　inductive　debris　detection　has　several　unique　advantages　such　as　simple　structure,　usability　for　metallic　flow　channel　and　non-sensitivity　to　the　quality　of　lubricant,　it　has　been　more　and　more　applied　in　the　mechanical　fault　diagnosis　and　condition-based　maintenance.　However,　the　inductive　principle　is　easily　interfered　by　electromagnetic　variations　and　mechanical　vibrations　in　the　actual　working　condition,　which　limits　the　detection　resolution　and　the　early　mechanical　wear　prediction.　Although　some　signal　processing　methods　have　been　proposed　to　improve　the　signal-to-noise　ratio　(SNR),　the　fundamental　denoising　idea　for　sensor　designs　is　still　absent.　This　paper　proposes　using　common-mode　rejection　to　eliminate　interferences　for　inductive　debris　detections.　The　mathematic　model　is　built　to　discuss　the　rejection　effect,　and　the　magnetic　distribution　is　simulated　to　analyze　the　sensor　structure.　Finally,　a　debris　sensor　is　designed　to　validate　the　performance　through　conducting　actual　debris　detections　and　external　interference　experiments.　The　results　indicate　that　the　sensor　can　effectively　eliminate　the　multiple　interferences　such　as　alternating　power,　oil　pulsation,　electromagnetic　variation　and　mechanical　vibration.　Especially,　the　SNR　has　been　improved　about　7.5　times　in　the　actual　debris　detection,　which　demonstrates　a　good　potential　for　applications.