Abstract: The SBM-Undesirable model based on non-expected output was used to measure and evaluate the eco-efficiency of grain production in 16 cities in Shandong Province from 2000 to 2019. ArcGIS visualization technology and Moran index were used to explore the characteristics of its spatial pattern, and a Tobit model was constructed to analyze its driving factors. The results showed that the eco-efficiency of grain production in Shandong Province showed a fluctuating upward trend, with an annual average value of 0.58, and there was still much room for improvement. Spatially, there were obvious differences, the areas with a high level of eco-efficiency of grain production had gradually shifted to the northwest and southwest of Shandong Province; except for 2002, 2006 and 2008, there was a significant positive spatial correlation, and the overall trend was “H-H”and “L-L”agglomeration. The level of financial support for agriculture, the income level of rural residents and the cropping structure had significant driving effects on the eco-efficiency of grain production in Shandong Province, and there were spillover effects on the eco-efficiency of grain production between different regions.

Key words: eco-efficiency of grain, SBM-Undesirable model, spatial-temporal patterns, driving factors, Shandong Province