A new reliability optimization allocation for multifunction systems considering differentiation of functions based on GO methodology is proposed in this paper. First, constraints considering differentiations of functions are proposed based on GO method, which are function importance factor constraint, and system reliability constraint, respectively. Then, the objective function of optimization allocation problem is built to minimize the system cost. Based on above, the mathematic model of reliability optimization allocation problem for multifunction systems considering differentiations of functions is established. In addition, an improved Ant Colony Optimization (ACO) is proposed to solve this mathematic model. Furthermore, the process of the new method is formulated. Finally, the new method is applied in reliability optimization allocation of Power-Shift Steering Transmission whose goal is to minimize the system cost. Compared with the results by using basic ACO, it is shown that the new method is reasonable, advantageous, and feasible for the reliability optimization allocation problem with differentiation of functions. Clearly, this study solves the disadvantages of the existing reliability optimization allocation methods efficiently so that it can quickly, efficiently, and directly allocate the system reliability index to design units for complex systems. All in all, this paper not only provides a new approach to conduct reliability optimization allocation for multifunction systems considering differentiation of functions, but also improves the theory and widens the application of GO methodology. In addition, this paper can also provide guidance for the similar reliability optimization problem.