Abstract: Based on the regional diversity characteristics of Yunnan, the entropy method was used to measure the agricultural green development level of Yunnan according to the states and cities, then ArcGIS 10.8 was used to divide five different levels of agricultural green development levels. Finally, the differences of agricultural green development paths in different levels of regions were deeply analyzed. The study found that, although the level of agricultural green development in different states and cities was constantly improving, the problems were different: (1) Some factors affecting the level of green development were the excessively high intensity of chemical fertilizers and pesticides, and the low efficiency of resource utilization, which was different from the general conclusion that the input intensity was too high. (2) The highest correlation of agricultural green development was environmental benefit, the second was resource benefit, and the third was output benefit, which showed that the focus of current agricultural green development was still in agricultural reduction of chemical fertilizer, pesticide and so on. (3) The correlation between agricultural green development level and each element was the output value per unit sown area, the agricultural electricity consumption index, the disposable income of rural residents, the number of three products and one standard product, the irrigation intensity of cultivated land and the intensity of agricultural film use.