Abstract: A fuzzy evolutionary pole placement (FEPP) approach is proposed for the two-wheeled self-balancing robot (TWSBR) system. The dynamical model of the TWSBR is firstly established based on Lagrangian formulation. The tangent linearization of the model at its equilibrium yields to the linearization nominal model. For the sixth-order underactuated system, the FEPP controllers are synthesized to achieve dynamic balancing and motion control of the TWSBR. Simulation and experiment results verify the effectiveness of the FEPP control strategy in dealing with underactuated, unstable and multi-variable coupled systems.