Abstract: Abstract：Exploring regional inequality, spatial-temporal pattern and dynamic evolution of carbon emission intensity from agriculture has important theoretical and practical significance for formulating a reasonable agricultural carbon emission reduction policy.In this paper, 31 provinces (municipalities and autonomous regions) in China were taken as the research unit to measure carbon emission intensity from agriculture during 1997-2016, through an integration of Theil index and exploratory spatial data analysis (ESDA) and R/S, we examine the regional inequality, spatio-temporal pattern and dynamic evolution characteristics of carbon emission intensity from agriculture.The results are shown as follows: (1） The carbon emission intensity from agriculture in China and the three major grain functional zones showed a downward trend, and the spatial distribution of different types carbon emission intensity from agriculture in province was larger, the high-intensity provinces of carbon emissions remained stable, while the low-intensity provinces are expanding; （2) The regional inequality of carbon emission intensity from agriculture presented an extending trend in 1997-2016, Among this, the agricultural carbon intensity inequality between regions is not obvious, but that within a region presented an expanding trend, and the growing rate of inequalities in the grain balance area tends to rise, the main production area rises first and then falls, and the main sales area tends to converge; （3) By observing the change of Moran’s I in 1997-2016, we found that the spatial agglomeration range of carbon emission intensity presented a narrowing trend, hot spots remain unchanged, but cold spots presented a shrinking tendency, Qinghai and Xizang formed a stable cold spot, and Beijing, Tianjin, Hebei, Shannxi, Shanxi, Henan and Shandong formed a stable hot spot; (4） The performance of carbon emission intensity from agriculture in China, the three major functional areas and the provinces had an obvious fractal characteristic, and carbon emission intensity value will show an evolutionary trend that continues to decline in future, especially with the downward trend of the grain sales area.