Abstract: Based on the panel data of 9 provinces （regions） in the Yellow River Basin from 2006 to 2019, the driving factors of agricultural water use were decomposed into technical effect, economic effect, urbanization effect, and population effect through the Kaya identity, and then the LMDI model was used to explore the impact of each effect on agricultural water use. Finally, the Tapio decoupling elasticity index was introduced to further explore the response relationship between economic indicators and agricultural water use. The results showed that, on the whole, the economic effect and population effect promoted the growth of agricultural water consumption, and the economic effect was dominant; the technological effect and urbanization effect inhibited the growth of agricultural water consumption, and the technical effect was dominant. From a regional perspective, the water consumption of Qinghai, Gansu, Ningxia, Inner Mongolia, Shaanxi and Shandong had declined, and agricultural water use had been effectively restrained on the whole, but Sichuan, Shanxi and Henan had not been effectively restrained. The nine provinces （regions） in the Yellow River Basin had basically achieved decoupling between agricultural water use and economic growth. Among them, Qinghai, Gansu, Ningxia, Inner Mongolia, Shaanxi and Shandong had strong decoupling, and Sichuan, Shanxi and Henan had weak decoupling.

Key words: agricultural water use, LMDI, decoupling, driving factor, Yellow River Basin