Abstract: An accurate prediction of the net irrigation water demand can greatly contribute to promoting the efficient use of water resources in the irrigation areas. Empirical formulas are normally used to calculate the effective precipitation of different crops for the estimation of the net irrigation water demand. However, the excess precipitation may cause deep seepage and surface runoff, leading to incorrect prediction. In this study, a calculation model of net irrigation water demand was established using the three states of soil moisture in the root zone, including the water-deficient, residual water, and full-water state. The model also fully considered the deep percolation and surface runoff caused by precipitation. Five independent parameters were evaluated for the net irrigation water demand, including the precipitation, crop water demand, root depth, field capacity, and wilting point. Taking the Jingdian Irrigation District in Gansu Province of China as an example, the annual average and monthly average net irrigation water demand, and the effective precipitation were calculated to clarify the influence of three variables (precipitation, crop water demand, and root depth) on the net irrigation water demand. The results showed that the annual net irrigation water requirement of different crops was between 319.4 and 732.3 mm, while, the precipitation use efficiency was between 39.2% and 56.1%. Furthermore, the precipitation use efficiency of summer-autumn crops was higher than that of spring-summer ones. The annual net irrigation water demand of summer-autumn crops presented a stronger correlation with the annual precipitation, whereas, that of spring-summer crops shared a stronger correlation with the crop water demand. Monthly net irrigation water demand for all crops was significantly correlated with the monthly crop water demand only. The crop water demand was positively correlated with the net irrigation water demand, while the precipitation and root depth were negatively correlated with the net irrigation water demand. The sensitivity of crop water demand to annual net irrigation water demand was between 10.62% and 13.49% for the nine main crops, the sensitivity of precipitation to annual net irrigation water demand was between 1.00% and 4.76%, and the sensitivity coefficient of root depth to the most crops was 0, where the soybean and wolfberry showed the minor effects. Summer-autumn crops were more sensitive to the precipitation and crop water demand than the spring-summer crops, where all crops were less sensitive to the root depth. Overall, the variables contributing to the net irrigation water demand were ranked in descending order of the crop water demand, precipitation, and root depth, of which 86.0% of crop water demand was absolutely dominant. Specifically, the contribution rate of precipitation to the net irrigation water demand was between 8.1% and 25.9%, and that of crop water demand to the net irrigation water demand was between 73.5% and 91.4%. The contribution rate of precipitation to the net irrigation water requirement for the summer-autumn crops was greater than 15%, the contribution rate of spring-summer crops to the net irrigation water demand was greater than 85%, and the average contribution rate of root depth was only 0.2%. The contribution rate of root depth to the net irrigation water demand reached 12.0% from 2000 to 2020, indicating that the root depth posed a significant impact on the net irrigation water demand in specific years. In addition to annual precipitation, the temporal distribution of precipitation also posed the contribution of root depth to the net irrigation water demand during the year. Consequently, the net irrigation water demand model can be expected to fully consider the difference in the effective precipitation of different crops, compared with the traditional one. The finding can provide a strong reference for the water resource management in the irrigation areas.