Inertial Measurement Units (IMU) are in highlight for joint and motion monitoring applications. Several IMU sensor fusion algorithms have been proposed in literature. Kalman Filter and its variants are the most used for more precision. However, they are computationally expensive. However, for faster computations, researchers and industry use complementary filter. More recently, a new variant of Kalman Filter was introduced as a Double Stage Kalman Filter in order to reduce the Kalman Filters computation cost. Our research investigates the performance of the Complementary and Double Stage Kalman filters in monitoring of joints in serial manipulators using Microelectromechanical-system MEMS based IMU. This study carried dynamic experiments using a serial robot to estimate the orientation of IMU, thus the joint angle of the associated segment. The study showed that both filters yield accurate estimations. The study showed also that Double Stage Kalman Filter has lower RMSE and achieves more precise estimates than Complementary filter mainly when the movement is around IMU x-and y-axis. Our findings indicate that the Double Stage Kalman Filter can achieve higher precision than the complementary filter using lower computation time than the former variants of the Kalman Filters in serial manipulator joint monitoring applications.