In this paper, a gas turbine bearing wear monitoring method based on the magnetic plug inductance sensor is presented. Using the induced magnetic field of the magnetic pulse generated by the inductance coil, this method is applied to measure ferromagnetic wear debris in the oil, and the condition of bearing wear can be monitored and predicted on line. To estimate the mass of different particle size of ferromagnetic debris, the sample databases of debris mass were built, the oil capturing experiment was conducted, and the mapping model for the voltage signal of the sensor and the captured accumulation of ferromagnetic wear debris based on the BP (Error Back Propagation) neural network was established. Moreover, the kernel method was used to calculate the voltage distribution of the debris sensor in the step time of wear prediction, and the mean value and confidence boundary value of the signals in the expected step time were obtained. Moreover, the prediction model of the bearing wear was established by a linear regression method to predict the mass of ferromagnetic wear debris generated by bearing wear. Finally, a lubricating oil debris detection system was designed, and the bearing wear test was conducted on the bearing wear testing rig. The results showed that the monitoring method can continuously monitor and dynamically predict the condition of bearing wear online with the advantages of stability and automatic purifying lubricant oil.