The femoral nerve innervates the quadriceps muscles and its dermatome supplies anteromedial thigh and medial foot. Paralysis of the quadriceps muscles due to the injury of the femoral nerve results in disability of the knee joint extension and loss of sensory of the thigh. A child could walk independently even though he had injured his femoral nerve severely due to the penetrating wound in the medial thigh. We measured and analyzed his gait performance in order to find the mechanisms that enabled him to walk independently. The child was eleven-year-old boy and he could not extend his knee voluntarily at all during a month after the injury. His gait analysis was performed five times (GA1~GA5) for sixteen months. His temporal-spatial parameters were not significantly different after the GA2 or GA3 test, and significant asymmetry was not observed except the single support time in GA1 results. The Lower limb joint angles in affected side had large differences in GA1 compared with the normal normative patterns. There were little knee joint flexion and extension motion during the stance phase in GA1 The maximum ankle plantar/dorsi flexion angles and the maximum knee extension angles were different from the normal values in the sound side. Asymmetries of the joint angles were analyzed by using the peak values. Significant asymmetries were found in GA1with seven parameters (ankle: peak planter flexion angle in stance phase, range of motion; ROM, knee: peak flexion angles during both stance and swing phase, ROM, hip: peak extension angle, ROM) while only two parameters (maximum hip extension angle and ROM of hip joint) had significant differences in GA5. The mid-stance valleys were not observed in both right and left sides of vertical ground reaction force (GRF) in the GA1, GA2. The loading response peak was far larger than the terminal stance peak of vertical ground reaction curve in the affected side of the GA3, GA4, GA5. The measured joint moment curves of the GA1, GA2, GA3 had large deviations and all of kinetic results had differences with the normal patterns. EMG signals described an absence of the rectus femoris muscle activity in the GA1 and GA2 (affected side). The EMG signals were detected in the GA3 and GA4 but their patterns were not normal yet, then their normal patterns were detected in the GA5. Through these following gait analysis of a child who had selective injuries on the knee extensor muscles, we could verify the actual functions of the knee extensor muscles during gait, and we also could observe his recovery and asymmetry with quantitative data during his rehabilitation.