Estimation of walking behavior such as gait pattern, speed and distance of walking, is discussed in this paper. In gait rehabilitation, measurement of the temporal and spatial parameters during gait is of critical importance for monitoring functional recovery. The 3-axis accelerometer on the waist belt is used to detect acceleration of body movement. From the filtered signal, we can estimate information about the subject such as gait pattern, speed of the subject, total walking distance, and improvement of walking function. Speed and distance are estimated by using step length. We performed on two healthy subjects to show the practical aspect of the proposed algorithm. The proposed scheme is simple to implement and it can be applied as a virtual controller to human-machine interfaced system.