In the light of the data on soil salt regime in 60 points in the northern part of Quzhou Couaty (247.4 km²), Huang-Huai-Hai plain, observed in spring and summer, 1987, and the study results of some predecessors. all kinds of factors influencing the salt regime in sails, such as the texture of soil profile (mainly the thickness of clay layer), the depth and saly content of groundwater, precipitation, irrigation and the initial saly content in soils are used as independent variables in the statistical models for prediuing soil salt content in spring and summer developed with the method of stepwise regression. The dependent variables are contents of total dissolved saly in 0-100 and 0-40 cm deep soils at next time interval. The factors seriously affecting the salt regime are all selected into the predicting equations. The independent variables thst,ase selected into the equations are different with the season and soil depth. There is a good agreement between the predicted results and observed data. Preliminary test demonstrates that the models can at least be used for the prediction of soil salinity in agriculture production. The prediuing models can only be used in such situations that the year type of predicting ime it the same as 1987 and the natural conditions are similar to Quzhou County's.