Sequence　stratigraphy　is　a　great　contribution　to　the　analysis　and　characterization　of　oil　reservoirs.　But　how　to　divide　sequence　stratigraphic　units　quantitatively　is　the　urgent　problem　to　solve.　Well　logging　signal　representing　lithology　and　physical　properties　embrace　much　information　related　to　sedimentary　cycles,　and　wavelet　analysis　of　signal　has　good　time-frequency　local　adjustable　performance　to　be　able　to　identified　different　frequency　cycles.　So　we　can　draw　the　conclusion　that　the　sediment　cycles　in　different　periods　can　be　identified　by　using　wavelet　and　analysis　of　well　log　curves.　In　terms　of　the　properties　of　the　wavelet　bases　and　the　characteristics　of　logging　signals,　dmey　12　is　chosen　for　the　cycle　classification　of　logging　curve.　The　research　shows　that　the　wavelet　analysis　is　a　helpful　complementary　technique　for　the　location　of　stratigraphic　sequences.