Abstract: Taking the Pinghu formation of Eocene in Well Pb?1 of Pingbei section in Baojiao slope belt in Xihu sag of Donghai basin as an example, the geologic information contained in GR logging data from Well Pb?1 Pinghu formation are analyzed using Matlab wavelet. Based on the time?frequency energy and curve characteristics of the wavelet transform coefficient after transforming, the coal?bearing strata of the Pinghu formation are divided into 3 long?term cycles and 5 middle?term cycles, which show that the wavelet transform has higher resolution in well logging sequence stratigraphic division, by which the characteristics of hidden sedimentary cycles can be recognized. On the basis of these, this paper discussed the relationship between the curves of wavelet transform coeficients in different scales and the coal measures positions. The result indicates that the coal measures in the study area are mainly located in the positions with relatively small peak value on the curves, which correspond to the delta retrograding process within the long?term cycles and the delta retrograding?prograding conversion period within the middle cycles