Power MOSFET plays an important role in power electronic systems, the failure of which will lead to system losing functions. Thus the degradation failure process and mechanism of MOSFET attract wide attention of scholars. However current research works cannot solve the real time degradation monitoring problem of MOSFET. Normally, MOSFET should be removed from the system for testing or monitoring. This paper presents a new in-situ prognosis method for MOSFET based on miller effect. According to the theory analysis, simulation and experiment results, the miller platform voltage is identified as a new degradation precursor. Then, particle filter algorithm is used for MOSFET remaining useful life (RUL) prediction based on the degradation data of miller platform voltage. Meanwhile, MOSFET samples with different degradation levels are formed by “overstress of gate bias” accelerating life test. The samples are placed into application circuit for the monitoring and extraction of miller platform voltage, so as to verify the method proposed in this paper.