Maximising the recovered regenerative braking energy during the deceleration can significantly reduce the Electric Vehicle (EV) energy consumption and increase the range. Compared with the Front Wheel Drive (FWD) or Rear Wheel Drive (RWD) EV, an All Wheel Drive (AWD) EV with 2 electric machines (e-machines) has more control degree freedom when developing the regenerative braking control strategy. By implementing the regenerative braking at the front axle, rear axle, or at the front and rear axles simultaneously, the amount of recovered kinetic energy will be affected. Furthermore, the e-machines at the front and rear axle in the AWD EV can have different sizes or be the same. Therefore, the ratio between front and rear e-machine power rating should also be investigated to understand its effect on the amount of recovered energy during deceleration. This paper starts with the analysis of the vehicle braking behaviour compared over different driving cycles, and the comparison of two configurations of regenerative braking system, Category A and B. Then, the AWD EV is modelled, and its regenerative braking controller is developed using Ricardo in-house, proprietary simulation tools. The power rating of front and rear axle e-machines in this model is varied. The regenerative braking controller simulates Category A or B regenerative braking system with various control strategies (such as front axle or rear axle only regenerative braking, and all wheel regenerative braking). Simulation is done to investigate: 1) the difference in recovered energy by implementing the regenerative braking at different axles with Category A or B systems, and 2) how the ratio between the front and rear axle e-machine power rating affects the amount of recovered regenerative braking energy. This in turn affects the overall brake balance distribution and impacts upon vehicle stability. Finally, the simulation result is analysed and discussed.