Abstract: Simplex Methods are quite efficient, but not very robust. They normally converge into a small number of iterations, but often fall into local minima. On the other hand, Genetic Algorithms are very robust, but unacceptably expensive. On the basis of analysis of the advantage and disadvantage of Simplex Methods and Genetic Algorithms, we present a hybrid method named SPGA aimed at coupling the efficiency of Simplex Methods and the robustness of Genetic Algorithms for solving the groundwater inverse problem. In the paper, we introduce the operators of SPGA and apply it to the inverse simulation of a water source area. As shown by the calculation results, the SPGA has strong point of high precision and fast speed.