The well production cleavage is the important premise for the detailed development of multi-layer reservoirs. The accuracy of production cleavage is the key to improve ultimate recovery, and the stimulation measures at late stage of oilfield development. The normal KH cleavage methods often only consider oil layers‘ static parameter，without considering the influence of inter-layer interference and other related dynamic parameters, it is bound to produce huge errors, and it has been difficult to solve the problem of multi-layer reservoir production cleavage. Therefore, based on the conventional method of production cleavage,with use of the inverse distance weighted interpolation method，construct a new forecast model of production cleavage, and the gray correlation analysis method is used to quantitatively compare the influencing factors in the cleavage coefficient. The main factors affecting the cleavage coefficient are determined, and the production cleavage of the oil well is finally realized. The method is applied to the yield of the X block in the Z oilfield, and the precision is obviously improved compared with the conventional method, which is compliance with actual development characteristics, and it provides an important scientific basis and guiding role for the later development of the reservoir.