Electromyography (EMG) is a signal that measures human muscle movements, and it is used to grasp human movement intent in human-machine interaction and to evaluate movement. An EMG signal that has the advantage of intuitively recognizing human motion can be analyzed with various features in time domain and frequency domain. However, the EMG signal changes its characteristics when fatigue occurs in the muscles. Most studies have investigated the muscle fatigue resulting from isometric conditions. Our research team want to use EMG signal in walking situation. Therefore, we analyze the characteristics of muscle at the time of walking and judge the condition of muscle. In this paper, we analyzed MAV and MNF among EMG characteristics during walking and analyzed their characteristics in three environments: level walking, ramp descent, ramp ascent