The servo system of a permanent magnet (PM) motor should always maintain fine torque and speed responses. Accurate motor parameter identification is necessary for the PM motor servo system because the current control system is designed by considering the electrical parameters of the PM motor. However, the motor parameters vary with the age of the motor and the temperature. Moreover, the current sensors have offset values. When the current sensor has offset values, the PM motor servo system produces torque ripple. In order to overcome these problems, this paper proposes a new real-time estimation method for both the current sensor offsets and the electrical parameters (resistance R a, inductance L a, and magnetic flux φ fa) of the surface permanent magnet (SPM) motor. The proposed method involves the use of a real-time algorithm and a current simulator, which is operated using a DSP software system. In order to accurately estimate the motor parameters, the proposed method uses estimated currents, the DC components of sensor currents, and the nominal motor parameter values. The experimental results confirm that the proposed method satisfactorily estimates the current sensor offset of the U phase and V phase, as well as the electrical motor parameters R a,L a, and φ fa accurately. © 2012 Wiley Periodicals, Inc.