Abstract: Objective:A strain with a high yield of γ-polyglutamic acid(γ-PGA)was identified,and its fermentation medium was optimized. Methods:Bacillus sp. N-2,a strain previous screened by mutation in our laboratory,was chosen as the starting strain,and then it was identified by its 16S rDNA sequence similarity analysis. In order to achieve a higher production of γ-PGA,the fermentation medium of the strain was optimized by single factor experiment and RSM(response surface method). Results:The result of 16s rDNA sequence analysis showed that the strain N-2 was Bacillus subtilis. In fermentation experiment,three factors(glucose,sodium glutamate,K₂HPO₄·3H₂O)that significantly affect the γ-PGA production of the strain were screened out using plackett burman test and their greatest response area were estimated by the design of steepest ascent approach;Futher,the extreme point of response surface was obtained by box-behnken design. The optimal conditions were as follows:glucose 42.93 g/L,sodium glutamate 44.85 g/L,K₂HPO₄·3H₂O 2.39 g/L. In this condition,the final yield of γ-PGA was 28.51 g/L after 54 h fermentation,which was 34.48% higher than before optimization. Conclusion:It needs less test using RSM,which can quickly optimize the fermentation medium components,get reliable results,and it is an effective way to increase PGA yield.