One wireless attitude measurement system based on a MEMS multi-sensor has been developed. The data from 3-axis MEMS accelerators and 3-axis MEMS gyroscopes are used to determine the pitch angle and the roll angle by using an extended Kalman filter (EKF) equation deduced from the direction cosine matrix. The attitude angles can be resolved from the EKF algorithm based on Quaternion. The system can be used in dynamic as well as static environments by adjusting the measurement noise matrix of the extended Kalman filter equation. Results indicate that system is determined with standard deviations below 5º even in high dynamics and 1º in static state. The system can collect wireless data from over six nodes simultaneously with a sampling rate up to 100 Hz. The system node measuring 22x15x10mm and weighing 4.5g contains wireless communication module, 3-axis gyroscope, magnetometer and accelerometer.