In this paper, a union estimation method for AHRS (Attitude and Heading Reference System) and sensor calibrations is proposed. Firstly, an iterated calibration model for accelerometers and magnetometers is proposed. Using the proposed iterated model, the scale factors errors, nonorthogonal errors, and bias errors are estimated in real-time. Meanwhile, the iterated model of misalignment angles of accelerometers and magnetometers is proposed. Then, the measurement errors of the accelerometers and the magnetometers are estimated. Using the estimated error parameters, the outputs of the accelerometers and magnetometers are calibrated. At last, using the calibrated outputs, the real-time attitude and heading reference system (AHRS) are constructed, all algorithms are processed in real-time. The simulation test is shown that the performance of the proposed method is superior to the same one of the traditional methods.