[Objective] The objective of this article is to reveal the spatial heterogeneity of the driving forces of per capita cultivated land area change in middle reaches of the Yangtze River based on geographically weighted regression(GWR) model in order to provide basis for the management of cultivated land resources in this area. [Methods] By investigating the current situation of cultivated land area per capita in this area, the Moran's I index of cultivated land area per capita was analyzed. A comparison was made between ordinary least squares(OLS) and GWR by using the relevant data. Based on GWR, a regression analysis on the influencing factors of cultivated land area per capita was analyzed in each city. [Results] (1) The influence of urbanization ratio on cultivated land area per capita varied from positive to negative correlations, with an enhancing influence degree and obvious spatial differences. (2) The growth rate of population and the per capita cultivated land area showed negative correlations in most areas, while they showed positive correlations in local areas, with a decreasing influence degree and large spatial difference. (3) In most areas, the proportion of the gross output of the first industry and cultivated land area per capita were positively correlated, while they were negative correlated in local areas, with a decreasing influence degree and large spatial difference. (4) The influence of grain yield per unit area on cultivated land area per capita changed from negative to positive correlations, with an enhancing influence degree. [Conclusion] The research shows that GWR is better than OLS in reflecting the spatial heterogeneity of the driving factors, and the results of GWR clearly reveal that different factors bring different degree of effect on cultivated land area per capita in different areas.