Abstract: It is important to quantitatively characterize biological nitrogen fixation (BNF) of soybean for farmland fertilizer management and non-point source pollution control. As a critical tool, BNF models play irreplaceable roles in predicting nitrogen fixation of soybean quantitatively. Domestic researchers have carried out some work on the influence of environmental factors on nitrogen fixation, however most of them are not from a quantitative point of view. BNF models can be classified into empirical-statistical models and process-based mechanistic models. In this paper, the two kinds of models are briefly introduced, and the 5 most commonly used mechanistic BNF models of soybean at the field scale abroad are highlighted. The response functions of different environmental factors to various models are compared, and the prospects for further BNF quantification research are discussed．