The information, which is mined from the drilling process monitoring signals data, could be helpful to inspect the holes drilling quality. This study presents a holes drilling consistency inspection method based on the fusion of monitoring signals marginal spectrum characteristics and bispectrum characteristics. At first, the three acceleration vibration sensor and acoustic emission sensor are used to monitor holes drilling process, and the Hilbert Huang transform and a high order spectrum estimation are used to analysis every hole drilling monitoring signals, so every hole’s the marginal spectrum and double spectrum features are extracted from the monitoring signals. Finally, the principal component analysis method is used to realize features dimension reduction, fusion features and features clustering, the computer conclusion show the change condition of these features directly and clear. Based on the coupling relationship between the drilling process quality fluctuation and the numerical changes of these features, and compared with the artificial quality test results of drilling hole, the result show that the data clustering research of the fusion of hole drilling monitoring signals marginal spectrum characteristics and bispectrum characteristics can realize the holes drilling quality consistency detection effectively, analysis and identify the abnormal drilling quality rapidly.