Taking the irrigated area of Liaofang water conservancy project in Jiangxi Province as an example, the main factors influencing the water demand were determined by referring to the yearbook, field investigation and expert consultation. The input samples of BP neural network were constructed according to the main influencing factors, and the Water demand model of irrigated area was programmed by Matlab software platform. The results show that BP neural network is an effective method to forecast the water demand of irrigation area in 2025 and BP neural network is successful in the application of water demand forecast in LiaoFang irrigated area.