In the context of the era when new quality productivity is promoting the construction of an agricultural power, sorting out the rising path of agricultural carbon in the Yangtze River Economic Belt and clarifying issues, such as agricultural carbon height and carbon decoupling are important to promoting the "double carbon" process in the agricultural field. A statistical model of agricultural carbon emissions was established to measure the agricultural carbon emissions of 11 provinces （municipalities） in the Yangtze River Economic Belt from 2000 to 2022. On the basis of clarifying the spatial and temporal distribution characteristics and reduction of agricultural carbon emission intensity, the LMDI model was used to analyze agricultural carbon emissions. Second, we constructed models of decoupling in terms of speed and quantity and explored the decoupling relationship between the most influential factors with the strongest driving effect and agricultural carbon emissions. The results showed that during the study period, the total agricultural carbon emissions in the Yangtze River Economic Belt first increased and then decreased, and the major emission sources were straw burning, grain planting, livestock breeding, and agricultural material input. The level of economic development was the major factor leading to the increase in agricultural carbon emissions. Although the improvement of agricultural production efficiency inhibited the increase in carbon emissions caused by the growth of output value to a certain extent, the total inhibitory effect was lesser than the total promotion effect. From the perspective of double decoupling, the region has not yet reached a stable and strong decoupling state between agricultural carbon emissions and economic income growth of farmers. Based on this, suggestions for pollution reduction and carbon sequestration, clean production, and green transformation are put forward, respectively, for the planting industry, breeding industry, and agricultural product supply chain in the upstream, midstream, and downstream regions.