Taking the grain-related data of 30 provinces in China from 2002 to 2020 as the research sample, based on the estimation of the flow of virtual cultivated land resources in grain trade, the Dagum Gini coefficient and its decomposition method, Kernel density estimation method and Markov chain analysis method were used to empirically investigate the regional differences and distribution dynamic evolution trend of virtual cultivated land resource flow in China's grain trade. The results show that: (1) the net import of virtual arable land resources in China's grain trade is on the rise, but the development is unbalanced between regions, showing the characteristics of "eastern> western > central"; (2) Inter-regional differences are the main source of regional differences in the net import of virtual arable land resources in China's grain trade, with a contribution share between 60%~70%, and the overall differences between the whole country and the three major regions are large, showing a trend of "rising first and then falling"; (3) The center development trend of net imports of virtual arable land resources in grain trade in the whole country and the three major regions gradually shifted to the right, except for the central region, the phenomenon of right tailing is significant, and the polarization phenomenon is alleviated; (4) When the spatial conditions are not considered, the net import of virtual cultivated land resources is more sustainable, and there will be no state transfer between provinces, but the trend to a high level is obvious. Therefore, it is necessary to clearly understand the spatial imbalance of the net import of virtual cultivated land resources in grain trade, and dynamically adjust the grain import and export trade policy, and at the same time pay attention to avoiding the negative impact of excessive import of virtual cultivated land resources in grain trade on the domestic market.

food trade  /  virtual cultivated land resources  /  regional differences