Using multi-factor statistical analysis, thousand kernel weights, brown rice rate, moisture, protein, fat, crude fiber, ash, carbohydrate and trace element content in millets planted in Shijiazhuang were detected and analyzed systemically, and a model for hierarchical comprehensive evaluation was established through principal component analysis. The results showed that influenced by heredity and environment, there was a significant positive relevance among minerals with significant differences among species; there was relatively small variation among fat, protein, crude fiber and carbohydrate; there was no significant correlation between thousand grain weights and protein content. Principal component analysis indicated fat, protein and ash content should be positively selected, while moisture, carbohydrate and crude fiber content should be negatively selected when establishing the evaluation model. Comprehensive evaluation using the established model showed that 11 varieties of millets could be classified into three categories, group 1 including varieties Chenggu13, 200131, Jigu19, Heng2001, Zheng066 and Bao213 showed the best quality and was most suitable to be planted in shijiazhuang, group 2 consisted of variety An074117, showed the worst quality, and group 3 including varieties 206058, An0474585, A2×Ce351and Cang372 showed moderate quality.