Abstract: A mathematical model is proposed for the parameter optimization problem in the multi-service CDMA network considering some network quality indices, such as the coverage rate, soft switching rate, service channel congestion rate, power consumption and so on. Through analyzing the characters of the proposed model, a constraint optimization genetic algorithm (COGA) is then designed and COGA's key technologies are given. Finally, the simulation results on a practical example show that the designed algorithm can achieve more effective configuration of network parameter than the practical DT (driver test) management does, which indicates that the proposed model and algorithm can address the parameter optimization problem in the multi-service CDMA network and then effectively guide the work of the practical CDMA management.