Abstract: In order to comprehensively and accurately evaluate the agronomic traits of summer soybean varieties（lines） in the Huang-Huai-Hai region（south）, and to construct a comprehensive evaluation system for soybean varieties（lines）, multivariate analysis methods such as factor analysis, DTOPSIS method, cluster analysis and regression analysis were used to analyze 15 characters of 37 varieties（lines） which accomplished the two-year regional trials in the Huang-Huai-Hai（south） summer soybean regional test from 2019 to 2021. The results showed that the variations of 15 agronomic characters of 37 soybean varieties（lines） was 2.4%-53.0%, the variation of mosaic virus disease was the largest, followed by numbers of effective branches, and the variation of growth period was the smallest.Ccorrelation analysis revealed complex correlations among the tested 15 agronomic traits, factor analysis transformed 15 traits into 5 main factors, with a cumulative contribution rate of 77.973%. The DTOPSIS method was used to calculate the comprehensive evaluation value（Ci value） of the 37 varieties（lines）, and the cluster analysis suggested 3 groups in 37 soybean varieties（lines）, each group had different dominant characters. Based on the Ci value, a comprehensive evaluation model of soybean varieties（lines） was established by stepwise regression analysis: Ci=0.484+0.033X₃+0.027X₄+0.067X₇+0.011X₁₁+0.034X₁₄,（R~2=0.955 0,F=132.027,P＜0.000 1）. This model can be used for the comprehensive evaluation of soybean varieties（lines） in the Huang-Huai-Hai region, thus to provide a reference for breeders to select high-yield and high-quality soybean varieties.