Abstract:  Synergistic efficiency of pollution and carbon reduction (SEPCR) is a core strategy for advancing ecological civilization. Accurately identifying its spatiotemporal characteristics and driving mechanisms is essential for formulating differentiated governance policies. This study investigates 284 prefecture-level cities in China from 2014 to 2023. A game-theoretic combined weighting method, coupling coordination model, and super-efficiency SBM model are integrated to measure SEPCR. Kernel density estimation and spatial Markov chain analysis are employed to explore its spatiotemporal evolution, while the MIC-RF-RFE hybrid model identifies key influencing factors. The XGBoost model and SHAP framework are further used to uncover the nonlinear relationships and spatial heterogeneity of these factors. Results show that: (1) From 2014 to 2023, China’s SEPCR exhibited a fluctuating upward trend, with the national mean rising from 0.2745 to 0.3039, and a spatial pattern of “East > Northeast > West > Central.” (2) SEPCR displays significant spatial dependence and path-locking effects: low- and medium-level cities have a 65.64% probability of maintaining their status with limited upward mobility, whereas high-level cities show strong stability with only a 24.82% probability of decline. (3) Population size, information infrastructure, and human capital are the dominant drivers, contributing 29.1%, 24.3%, and 21.5%, respectively. PS and HC show inverted U-shaped effects, while IDL exhibits an N-shaped relationship. (4) Regional heterogeneity is evident: SEPCR in the East is driven by innovation and human capital; in the Central region, constrained by resource endowment; in the West, limited by infrastructure and water resources; and in the Northeast, hindered by low innovation conversion efficiency. These findings suggest that optimizing factor allocation according to regional characteristics and strengthening cross-regional governance and policy coordination are crucial for enhancing sustainable SEPCR.