Abstract: In order to improve the accuracy of a handwritten Chinese character recognition system, a model is presented, which combines SCR and post-processing system. Genetic algorithm is applied to optimize system parameters. Tested on an online handwritten Chinese character recognition system（SCR）, 87.59% accuracy is achieved when first character candidate accuracy rate is 69.46%, and optimization contributes 6.4% improvement. The experiment results show that genetic algorithm （GA） is an effective method to optimize system parameters.