The resource curse effect hinders the economic growth of resource abundant regions. With regard to the current situation of water shortage in China, based on the provincial panel data and referring to the resource curse hypothesis, the resource curse effect of the agricultural virtual water is empirically validated through preliminary tests and by establishing a regression model for dynamic panel data of time effect in 29 provinces (regions) of China from 1993 to 2012. The resource curse troubled provinces are determined by formulating a resource abundance-economic growth quadrant graph. The results show that the resource curse effect of agriculture virtual water objectively exists in the short term in China. However, with the elapse of time, the resource curse effect of agricultural virtual water tends to weaken.