A novel attitude estimator for applications under dynamic conditions is proposed in this paper, which is based on the vector observations of gravity, geomagnetic field, and angular velocity. After a brief review of the existing solutions for attitude estimation and external acceleration rejection, we adopt the deviation of attitude, gyro bias, and external acceleration to construct the state vector in the proposed algorithm. Numerical simulation results confirm the feasibility and reliability of the proposed method, and also reveal the impacts of parameter tuning under different circumstances.