Exploring the peak process of carbon emissions from crop production provides a basis for mitigating greenhouse gas emissions. Previous studies found that carbon emissions from crop production in China reached an inflection point in 2015. Nonetheless, determining whether a peak has been reached is unreliable without verifying the specific peaking process using statistical approaches. To better understand the peaking process, this study calculated the carbon emissions from crop production in 30 Chinese provinces from 2000 to 2020, considering four carbon sources: agricultural materials, rice paddies, soil management, and straw burning. The peak carbon emissions process was then explored at the national and provincial levels. The Tapio decoupling index was used to verify the relationship between carbon emissions and economic output. The results showed that: (1) The total carbon emissions from crop production in China had an annual average of 233.269 Mt, increasing from 200.020 Mt to 242.819 Mt during the study period, peaking at 262.648 Mt in 2015. The average annual rate of change after reaching the peak was −1.560%, indicating that emissions entered a plateau. Over time, agricultural materials became the primary emissions source (34.6%), whereas soil management contributed the least (11.6%) in 2020. (2) Carbon emissions from crop production were positively correlated with the cropping scale. Only two provinces, Hunan and Henan, had the highest emissions of over 20 Mt; five provinces, such as Hubei and Shandong, had the highest emissions distribution of 15–20 Mt, and other five provinces, like Jiangxi and Sichuan, had the highest emissions ranging from 10 to 15 Mt. In contrast, the highest emissions in 18 provinces were less than 10 Mt, especially in Beijing, Tianjin, and Qinghai, with emission peaks below 1 Mt. As far as the peaking process, the carbon emissions in 13 provinces, including Beijing and Tianjin, were in a state of decline, those of 10 provinces, such as Shanxi and Chongqing, entered a plateau, and those of seven provinces like Henan and Anhui had not met their peak yet. (3) At the national level, the long-term relationship between carbon emissions and economic output showed weak decoupling, whereas the short-term relationship changed from weak to strong decoupling. At the provincial level, the short-term relationship evolved from multi-type coexistence to strong decoupling. Consequently, it is recommended that the emission mitigation of crop production in China should be accelerated by source and phase based on the peaking process and emission magnitude. Provinces with emissions in peaking and plateauing states require additional attention, as their subsequent developments determined the overall emission reduction. In comparison, flexible space for emission mitigation can be provided to the provinces in declining states, as many of them are accompanied by low emissions and optimistic momentum. However, three high-emission provinces, Hubei, Jiangxi, and Shandong, also reached their peak emissions and began to decline, which may serve as examples of provinces with similar conditions. These findings provide local solutions for accelerating the peaking process of carbon emissions from crop production in China.