This paper proposed a novel quantum particle swam optimization algorithm based on multi agent system(MAQPSO) approach to the reactive power optimization.The algorithm overcome the defect of conventional quantum particle swam optimization slow convergent speed and easy convergenve to local minimum point of error function on later. This algorithm combined quantum particle swam optimization algorithm with multi-agent system platform.Every Agent serves as a particle of quantum particle swam,competition by the Agent of the neighborhood and selfstudy,the algorithm can more quickly and more accurately converge to global optimal solution.Based on IEEE14, 30, 57 and 118 nodes system optimization simulation, the results show that the algorithm has high convergence precision and speed of optimization.