This paper presents a new strategy of direct torque controller for bearingless induction motor using space vector pulse width modulation based on fuzzy adaptive control. when we use direct torque controller using space vector pulse width modulation to take decoupling, the parameters of PI controller which generating the reference voltage vector in conventional SVM-DTC are difficult to determine the dynamic operation. In order to improve away the disadvantages of conventional SVM-DTC system, flux and torque fuzzy adaptive controller were designed to substitute the original flux and torque PI controller in the controlling for bearingless induction motor using space vector pulse width modulation. With the fuzzy algorithm, it is easy to obtain the control voltage component of the flux and torque respectively. Two voltage vectors achieve real-time adjustment and solve the disturbance problems in torque loop and flux loop. In this paper, the design process of the fuzzy adaptive controller is given. Use Matlab/Simulink to check the improved and traditional SVM-DTC method. The results show that the improved algorithms have a better performance in reducing the ripple of torque, flux and Rotor displacement when compared with the tradition DTC method. And it also improves the system dynamic performance.