Abstract: The paper constructs the index system of agricultural power, calculates and evaluates the level and spatiotemporal evolution characteristics of agricultural power in 30 provinces (autonomous regions and municipalities) in China from 2010 to 2020 by the entropy TOPSIS method, non-parametric kernel density estimation (KDE) and Moran’s I index scatter diagram. The results show that: (1) From 2010 to 2020, the level of China’s agricultural power was on the upward trend, but regional development was not synchronized. The top three provinces in the average value were Shandong (0.471), Jiangsu (0.401) and Henan (0.399), which were the best in eastern region, followed by the central region, the northeast region and western region. (2) The development level of China’s agricultural power has certain geographical spatial agglomeration characteristics, the eastern region is mainly “high-high” combination, the western region is mainly “low-low” combination, the central and northeast regions are mainly “high-low” and “low-high” combination. (3) The products of agricultural power index is relatively good, the supply guarantee level is second, and the development of science and technology equipment and management system is slightly backward. The level of China’s regional agricultural powers is on the rise, and the spatial difference of development always exists, but the difference of time evolution is not obvious. China has a vast territory, and the endowment difference between different regions is very obvious. In view of this, it is proposed to formulate and implement the strategy of agricultural power according to local conditions, constantly improve the modern agricultural production and operation system, promote regional cooperation, and strengthen the coordinated development among regions.