Abstract: Carbonate reservoirs show fractures and strong heterogeneity, leading to limited applications of a single logging or seismic method in fracture identification. Therefore, this study proposed a method for predicting fractures in carbonate reservoirs based on a fracture identification-sensitive log-seismic parameter (FISLSP) model, and the details are as follows: (1) Based on petrophysical analysis results, as well as formation micro-imaging (FMI) logs and the interpretations of fractures at well locations, this study established a segmental fracture identification-sensitive parameter (FISP) model for fractures at well locations; (2) Using the optimal seismic attributes of near-well seismic traces determined based on the high-quality fractured reservoirs classified according to geological and log interpretation results, this study constructed a FISP model based on the seismic attributes of near-well seismic traces (FISPSA); (3) Using logs combined with seismic attributes, this study established the technology roadmap, method, and procedure for reservoir fracture prediction based on a FISLSP model, achieved the 3D seismic prediction of fractures, and developed a FISLSP model-based 3D seismic fracture prediction method, which applies to buried-hill carbonate oil reservoirs. This method was applied to the prediction of fractures in the Chengdao buried-hill reservoirs of the Shengli Oilfield in the Bohai Bay Basin. The predicted results were consistent with fractures revealed through drilling, indicating that the new method is feasible. The results of this study warrant wide application to the fracture detection of similar buried-hill reservoirs.