[英文摘要]： Using a crop-water-salinity production function and a soil-water-salinity dynamic model, optimal irrigation scheduling was developed to maximize net return per irrigated area. Plot and field experiments were used to obtain the crop water sensitivity index, the salinity sensitivity index, and other parameters. Using data collected during 35 years to calculate the 10-day mean precipitation and evaporation, the variation in soil salinity concentrations and in the yields of winter wheat and cotton were simulated for 49 irrigation scheduling that were combined from 7 irrigation schemes over 3 irrigation dates and 7 salinity concentrations of saline irrigation water （fresh water and 6 levels of saline water）. Comparison of predicted results with irrigation data obtained from a large area of the field showed that the model was valid and reliable. Based on the analysis of the investment cost of the irrigation that employed deep tube wells or shallow tube wells, a saline water irrigation schedule and a corresponding strategy for groundwater development and utilization were proposed. For wheat or cotton, if the salinity concentration was higher than 7.0 g L^-1 in groundwater, irrigation was needed with only fresh water; if about 5.0 g L^-1, irrigation was required twice with fresh water and once with saline water; and if not higher than 3.0 g L^-1, irrigation could be solely with saline water. ; project supported by the National Natural Science Foundation of China （Nos. 50339030 and 90202001）.