To prevent falls in the elderly, gait measurements such as several-meters walking test and gait trainings are carried out in community health activities. To evaluate the risk of falling of the participant, it is necessary to measure foot contact times and positions so that the stride length of each leg and the walking speed can be used as evaluation parameters. However, the conventional measurement systems are difficult to install for use in community health activities because of their scale, cost and constraints of the measurement range. In this study, we propose a novel gait measurement system which uses an autonomous mobile robot with laser range sensor (LRS) for a long-distance walking test in a real living space regardless of detection range of sensor. The robot sequentially estimates its own pose and acquires the position of both legs of the participant. The robot leads the participant from the start to the goal of the walking test while maintaining a certain distance from the participant. Then, the foot contact times and the positions are calculated by analyzing estimated position and speed of each leg. From the experimental results, it was confirmed that the proposed robot could acquire the foot contact times and positions.