Abstract: It is the important research direction to realize intelligent optimization design in underground mining.so, this paper present a method to optimize stope parameters based on the combining of evolutionary algorithm, orthogonal test and FLAC3D.Firstly, three dimension model of stope with different parameters are established by orthogonal-simulation, then, genetic programming is used to obtain the nonlinear relationship between displacement index and stope parameters, finally, the optimization parameters are gained by means of genetic algorithm. Engineering application shows that this method is a fast optimization of stope parameter, which has good practical value.