Switching control is a key technology in a paralleled fuzzy neural network controller composed of fuzzy sub-controller and neural network PID sub-controller. Incorrect switching control will seriously affect the stability and rapidity of control systems. Fuzzy soft handoff control takes the fuzzy quantities of system error and error change rate as a basis for switching two control modes. The related weights of two control algorithms are continuously adjusted in the switching process. When one is falling, the other is rising. Smooth replacement between two different control modes can be realized. Simulation results for motor show the control system with fuzzy soft handoff possess battle characteristics of dynamic and steady state.