The aerobic endurance is considered an important physiological capacity of soccer players which is examined by Incremental Exercise Test (IET). However, it is not clear how general fatigue induced by IET influences physiological and biomechanical gait features in soccer players and how players recover optimally at post-IET. Here, the effect of general fatigue induced by IET on energy cost, gait variability and stability in soccer players was investigated. To identify an optimal recovery mode, the effect of walking at Preferred Walking Speed (PWS), running at Individual Ventilation Threshold (IVT) (two active recovery modes), and Rest (a passive recovery mode) on aforementioned features were studied. Nine male players walked 4-min at PWS on a treadmill prior IET (PreT), which was followed by four 4-min walking trials (PosT-0, 1, 2, and 3) with three 4-min recovery intervals (PWS, IVT, or Rest) between them, in three sessions (one for each recovery mode) in a random order. Energy cost, gait variability and stability were examined at PreT (baseline), and at PosT-0, 1, 2, and 3 (intervals of respectively 0–4, 8–12, 16–20, 24–28 min at post-IET). Gait variability was assessed by the standard deviation of trunk angle and gait stability was assessed by the local dynamic stability of trunk angular velocity. Gait stability was not affected by IET, despite increases in gait variability and energy cost. Different from IVT, PWS and Rest recovery modes reduced energy cost at post-IET. Gait variability and energy cost recovered at PosT-1 and PosT-2, suggesting that 8–12 and 16–20 min recovery intervals, respectively, were required for returning to their baselines. No preference for active over passive recovery was found in terms of gait variability and energy cost.