Abstract-A wearable robot is constantly in contact with its user. To measure and control the human-robot interaction force is important to properly and safely perform tasks together with the wearer, such as walking and load carrying. To avoid the burdensome integration of force sensors at the humanrobot attachments, we propose the use of a Kalman filter to perform data and sensor fusion to estimate the interaction force. We demonstrate the impact of real world issues on the measurement estimation and evaluate its performance in hardware. Preliminary simulation and experimental results demonstrate that good measurement prediction and control performance may be achieved with such an estimator.