Analysis of the spatiotemporal evolution of agricultural production efficiency and its determinants in China

China faces regional disparities in agricultural development. Understanding the spatiotemporal evolution of agricultural production efficiency and its driving factors is essential for advancing sustainable agricultural development. Using panel data from China’s 31 provinces from 2013 to 2023, we constructed an input–output indicator system and applied the super-efficiency slacks-based measure (SBM) model to evaluate agricultural production efficiency. The results indicate that the evolution of agricultural production efficiency went through three stages. The period from 2013 to 2016 was characterized by fluctuations; the period from 2016 to 2021 was characterized by an overall upward trend; the period from 2021 to 2023 was characterized by a downward trend. The number of provinces with an agricultural production efficiency score of ≥1 increased from 8 to 11. Temporal kernel density analysis revealed that polarization between high- and low-efficiency provinces weakened from 2013 to 2022 but intensified in 2023. Spatial kernel density analysis showed a decline in the spatial clustering of agricultural production efficiency, with high-density zones gradually expanding from coastal areas to inland regions. Geographic detector analysis identifies socioeconomic factors (per capita education expenditure, farmer income, agricultural development level) and agricultural structure (pesticide input, planting structure) as key determinants, with rankings showing moderate sensitivity to temporal aggregation and discretization method but substantive policy priorities remaining consistent. Region-specific issues, including the low comparative returns from agriculture in the eastern region and low production efficiency in the central grain-producing region, underscore the need for countermeasures tailored to regional patterns of agricultural production efficiency across different regions. Such measures could include increased investment in human capital and optimization of crop structures. These findings provide empirical support for region-specific strategies aimed at improving agricultural production efficiency and promoting sustainable agricultural development.