Dairy farming has become one of the most important industries for China’s national economic development, but the dairy farming process emits large amounts of greenhouse gases, which accelerate the greenhouse effect. Improving green production efficiency is important for the harmonious development between dairy farming and environmental protection. This paper explores the evolutionary characteristics of China’s GMLMI (Green total factor productivity of dairy farming) in terms of scale heterogeneity based on data from 2007 to 2018 from 27 major dairy-producing provinces (municipalities directly under the central government) in China, considering negative output indicators and using the super-efficient SBM-GML index model. The results show that 1) China’s GMLMI in 2007–2018 is highest in the medium scale, followed by the large scale, and lowest in the small scale. The regional distribution is highest in the central region, followed by the eastern region, and lowest in the western region. 2) China’s GMLMI shows a positive increasing trend from 2007 to 2018, and the decomposition indicators of the three scales (small, medium, and large) perform differently, indicating that there are differences in the ways to enhance the GMLMI of the three scales. 3) The results of the meta Frontier surface are different from those of the group Frontier surface, with the meta Frontier surface GMLMI being lower and the group Frontier surface GMLMI being overestimated. Finally, this paper proposes policy recommendations on how to improve the overall GMLMI of dairy cattle environmental farming in China.