Abstract: A novel hybrid simplex method and particle swarm optimization(HSMPSO) algorithm is presented.Comparison experiments with other published methods on 10 benchmark functions are conducted,and the extensive(analysis) on the effect of different parameters on the algorithm is given.Experimental results indicate that the hybrid method can improve both solution quality and success rates on most selected test functions compared with other alternatives,especially on multimodal functions optimization.Although very easy to implement,this hybrid SM-PSO is an efficient way to locate global optima of continuous multimodal functions.