In order to practice the concept of‘lucid waters and lush mountains are invaluable assets’ and promote the green development of agriculture, it is necessary to improve the efficiency of agricultural energy utilization. Based on the panel data of 28 provinces from 1995 to 2018, this paper calculated China's agricultural energy input from two categories of direct energy and indirect energy, and used EBM (Epsilon-based Measure) mixed distance function model to measure the energy efficiency of agriculture in China. The nuclear density function and spatial autocorrelation were used to analyze the dynamic evolution of agricultural energy efficiency, and the dynamic panel model was used to analyze the influencing factors of agricultural energy efficiency. The results showed that: ① From 1995 to 2018, the total agricultural energy input had increased year by year in China, with an average annual growth rate of 2%. Energy input structure changed from indirect energy-based to direct energy-based. Agricultural energy efficiency showed an evolutionary trend of ‘rising-stagnating-rising rapidly'in China. The agricultural energy efficiency was generally low in China, and there was a large space for improvement in agricultural energy efficiency. ② From 1995 to 2018, the average annual growth rate of agricultural energy efficiency in the eastern, central and western regions was 2. 7%, 1. 9% and 1. 4% respectively. In 2018, the agricultural energy efficiency in the eastern, central and western regions was 0. 81, 0. 71 and 0. 59 respectively. The gap between regions was expanding rapidly, and the agricultural energy efficiency in the central and western regions needed to be improved. ③ From 1995 to 2018, the agricultural energy efficiency of each province was polarized and the absolute gap was widened. There was obvious improvement in agricultural energy efficiency in Guangdong, Shandong, Jiangxi, Jiangsu, Liaoning and Tianjin, while the agricultural energy efficiency of Xinjiang, Guizhou, Zhejiang, Shanghai, and Inner Mongolia deteriorated. ④ From 1995 to 2018, there was no global spatial correlation of China's agricultural energy efficiency. However, local ‘high-high'concentration gradually appeared in the eastern region since 2010. ⑤ The first lag of energy efficiency had a significant positive impact on agricultural energy efficiency, and agricultural energy efficiency improvement had a time lag. The level of human capital, per capita net income of farmers and the level of urbanizaton had a significant positive impact on agricultural energy efficiency. The disaster rate, the level of development of secondary and tertiary industries, and the level of opening up had a significant negative impact on agricultural energy efficiency. In the implementation of the strategy of rural revitalization, we should focus on the central and western regions, take the cultivation of professional farmers as the key, focus on improving agricultural production conditions, enhance the level of cooperation between regions, exert the leading role of the secondary and tertiary industries, and enhance the ability of agricultural disaster prevention and mitigation. © The Authors