["Australian Economic Papers, EarlyView. ", "\nABSTRACT\nAgricultural carbon emissions are a key source of nonenergy greenhouse emissions in China, yet their spatiotemporal dynamics, regional disparities and structural changes remain insufficiently explored. This study quantifies agricultural carbon emissions in 31 provinces and municipalities in China from 2001 to 2020, and examines their spatiotemporal trends, regional disparities, and structural changes using the spatial Moran index to detect spatial dependence and clustering of agricultural emissions, and Theil index to decompose regional disparities. The results show that total agricultural carbon emissions rose from about 906 million tons in 2001 to a peak of approximately 1021 million tons in 2015, before declining to around 930 million tons in 2020. Spatial analysis with the global Moran's I index reveals a consistent, and at times significant, positive spatial auto‐correlation, indicating regional clustering of emissions. The Theil index showed a slight decrease over the study period, reflecting a narrowing of interregional disparities, particularly between eastern and northeastern provinces. Structural analysis identifies livestock and poultry as the largest contributors, accounting for about 44% of total emissions, followed by land‐use‐related emissions (about 38.5%) and methane emissions from rice cultivation. These results underscore both the spatial concentration and the changing composition of agricultural emissions, highlighting the need for spatial coordinated and sector targeted mitigation policies, especially in livestock‐intensive regions, to achieve emission reduction while maintaining agricultural output productivity in China's low carbon transition.\n"]