Abstract: Scenario planning and stochastic programming is introduced into the location selection of maritime rescue base and the optimization of rescue ship configuration. Several types of rescue ships and their respective seaworthy areas are investigated. Three severe weather scenarios are assumed. A two-stage stochastic programming model is built and the immune PSO（Particle Swarm Optimization） algorithm is developed. The method is verified with South China Sea case. The sensitivity analysis for the results is carried out. The result shows that the method can provide optimum proposals in saving rescue costs.