this paper reports the use of fiber Bragg grating (FBG) sensors for electric machines' in-service bearing condition monitoring. The proposed method enables simultaneous extraction of thermal and mechanical information on the bearing operational status for health monitoring and fault diagnosis purposes. The work first reports the instrumentation, calibration and the measurements interpretation methods of an FBG array sensor fitted to the drive-end bearing of an operating inverter driven induction motor. Experimental tests are then undertaken on a purpose build test rig to evaluate the proposed in-situ sensing scheme's performance under healthy and faulted bearing conditions. The results demonstrate that the measurements acquired by individual FBG sensors fitted in the bearing architecture contain concurrent thermal and mechanical information, which can be clearly differentiated to enable understanding of the examined motor bearing's thermal only and mechanical only operating conditions. Furthermore the in-situ measurements obtained in fault conditions are shown to contain clearly defined time and frequency domains fault signatures, which can enable unambiguous fault diagnosis and trending.