In polymerase chain reaction (PCR), in order to amplify massive DNA sequences successfully, it needs to design an appropriate primer pair. The constraints derived from the traits of PCR for proceeding PCR are used in searching for primer pairs. In this paper, in order to decrease the searching space and to increase the feasible quality of primers, a double orthogonal arrays intelligent crossover genetic algorithm (DOAIGA) is used to solve the primer design problem. DOAIGA combines the traditional genetic algorithm and the Taguchi methodology to efficiently search feasible primers under required constraints. The proposed intelligent crossover subsystem mainly concentrates on the better genes more systematic. The key point of DOAIGA is to achieve the elitism goal by applying the orthogonal arrays (OAs) that is used in quality engineering with a small amount of experiment features. In this thesis, the double orthogonal arrays are used to approach a better forward and reverse primers separately. Compared to the current existing softwares, DOAIGA can obtain feasible primer pairs more effectively. Finally the correctness of primer pair is verified by PCR experiment.