In our research, we focus on estimation of upper limb motions using skin deformation. Skin deformation is caused by the activities of various body tissues such as the muscles, tendons, and bones related to the motions. Especially, we can observe the activities of the deep layer muscles from the skin deformation. In this dissertation, we propose estimation methods of upper limb motion based on the skin deformation. Two types of the sensing devices are developed to measure the whole circumference of the skin deformation on the forearm and upper arm. To deal with various applications, we develop two categories of the motion estimation. In the first category, we develop a motion recognition method base on forearm deformation. The recognition method classifies seven different types of motions including the motions caused by the activities of deep layer muscles. In the second category, we develop two pose estimation methods. One method uses the forearm deformation and the other uses the upper-arm deformation. These pose estimation method estimates multiple joint angles of the upper limb. In the experiments of the motion recognition method based on the forearm deformation, we confirmed that our proposed method could recognize seven different types of hand motions including the motions related to the activities of not only surface layer muscles but also deep layer muscles. In the experiments of the pose estimation based on the forearm deformation, we confirmed that our proposed method could estimate multiple joint angles of the upper limb pose. In the experiments of the pose estimation based on the upper-arm deformation, we also confirmed that the forearm pose estimation is possible by using only upper arm deformation. These results of the experiments demonstrated that the skin deformation is useful for the upper limb motion estimation.