The conventional prediction model of rock drillability is based on regression analysis using well logging data. Regression analysis directly uses initial data to establish model, so its precision is not satisfied. By accumulating initial data, gray theory model (GM (0,N)) is able to weaken the random of initial data. Therefore, a practical approach to calculate the rock drillability, which base on GM(0,N) using well logging data, is presented in this paper. Based on the inherent relation of well logging information and rock drillability, a lot of logging parameters are selected closely to rock drillability, and the prediction model of rock drillability are established by GM(0,N). This method is applied to logging data process of Du4 well in SC oil field, the results show that it can improve predicting accuracy of rock drillability, and can easily frame rock drillability profile in some areas.