Abstract: A method for diagnosing the rolling bearing faults in a gearbox based on Adaptive Noise Canceling(ANC) and Demodulated Resonance Technique(DRT) was presented.The signals under normal condition were detected first,and then the continuously monitored signals of the rolling bearing were collected using the same sensor and measurement system at the same location of the gearbox.ANC processing was carried out using the signals under normal condition as the reference input and the monitored signals as the main input.Finally,DRT was applied to the noise-cancelled signals in order to realize feature extraction and fault diagnosis of the rolling bearing.The experimental results show that this method not only improves the signal-to-noise ratio of the measurements,but can also effectively diagnose the incipient faults of the rolling bearing in the gearbox.