For the problem of test selection for complex system, a test selection optimization based on Simulated Annealing Binary Particle Swarm Optimization (SA-BPSO) algorithm was adopted. The probabilistic jumping ability of simulated annealing algorithm was used to overcome the deficiencies of the particle swarm being easily fall into local optimal solution. The process and key steps of the algorithm for test selection in complex system were introduced, and the complexity of the algorithm was analyzed. The simulation results show that the algorithm has better performance in running time and testing cost compared to genetic algorithm, thus the algorithm can be used to optimize test points of complex system.