Abstract: Accurate fracture prediction plays a significant role in the exploration of carbonate reservoirs.However,it is difficult to identify fractures due to extremely complex anisotropic characteristics in fractured carbonate reservoirs.The paper integrates seismic technique with geology means to comprehensively quantitatively predict fracture.On the one hand,pre-stack P-wave anisotropic inversion is adopted to obtain fracture density and orientation.On the other hand,3D post-stack seismic attributes analysis is employed to detect relatively large-scale fracture.In addition,stress field simulation based on finite element analysis is used to discriminate single and multiple sets of fractures.Subsequently,the stress simulation result is used to constrain the results from seismic prediction to optimally characterize fractures.Finally,the method of combining the results from pre-stack and post-stack results with outcomes from stress analysis is applied for fracture prediction in ZG8 area of Tarim Basin,where fractures are comprehensively developed.The results are in good agreement with FMI information and geologic settings,which suggests that the combination of seismic detection method and stress field prediction is a feasible way to delineate fracture property in complex carbonate reservoirs.