Taking the water demand data from 1999to 2010of Nanjing for example, this paper analyzes the main factors that influence the water resource quantity based on the principle component analysis method. According to these main factors, the input samples of BP neutral network are determined. Thereby, the BP neutral networks can be trained to predict. The results show that population, GDP, water consumption of ten thousand yuan GDP, water resources per capita and volume of sewage discharge per year are the primary indexes that affect water resource demand. The corresponding prediction modeling outcome shows that the simulated experiment is quite fit for the practical situation and the average error of prediction is less than 0.2×108m3.