Motion measurement systems play an important role in a wide range of fields such as robot motion control and human motion analysis. Motion measurement methods using a camera, which is an external sensor, have problems such as low sampling rate and limited measurement range. On the other hand, the method using the encoder or inertial sensor, which is an internal sensor, has almost no limitation on the measurement range. Moreover, it can be measured at a high sampling rate. However, when using the internal sensor, it was necessary to use the kinematic model and kinematic parameters of robots and humans. Errors in these parameters lead to reduced accuracy in kinematic calculations. Therefore, the control performance and analysis accuracy are reduced. To solve these problems, we propose a method for estimating kinematic parameters using the inertial sensor. The proposed method uses a kinematic relational expression in the acceleration dimension. Therefore, kinematic parameters can be estimated without using absolute position information. In this paper, the proposed method is applied to the 3-link manipulator and the human body. The effectiveness of the proposed method is evaluated by comparing the estimated link length with the measured value.