Abstract: In order to identify the fertility, stability and adaptability of spring soybean in the regional variety test in the northeast of China, and at the same time to enrich the analysis methods of experimental results, the SHMM model was introduced to analyze the experimental result in the prematurity area of the springtime soybean variety in the North China from 2006 to 2007. Before principal component analysis using SHMM model, all the data was minused by the mean of the collectivity to get the eigenvalue and eigenvector, and then finally more dispasal sum can be achieved than linear regression madel and AMMI model. The result indicates that: the quality was high in the Qitai seed supervisal station, the Yanbian Agricultural Academy Institute, and the 71 corps of the agricultural fourth division in xinjiang province; the quality was low in the seed company of beilin district, the seed company of bayan county, the original seed station of dunhua city, and so on; some varieties have broad adaptability, such as Ha98-3964, Hai3007,Gongjiao99104A-1, Gang9775-2; He99-718 have especial adaptability in Yanbian Agricultural Acamedy Institute canton and Qitai seed supervise station; Jili soybean 2 has broad adaptability except for Qitai seed supervise station; Ha98-3964 and Hai3007 have high yield and stability; Gongjiao99104A-1 has low yield and stability; He99-718 has high yield but not stable.