Finding a green development trajectory that harmonizes economic growth with ecological conservation is essential for China's modernization process. Improving urban green development efficiency (UGDE) and attaining its convergence are among the potent strategies for fostering urban sustainability. Utilizing the super SBM model,this research quantified UGDE across 269 prefecture-level and above cities in China spanning from 2003 to 2021. Then,geographic information analysis and kernel density methods were applied to estimate the spatial patterns and temporal evolution of the UGDE. Moreover,the research examined efficiency disparities across diverse hierarchical levels employing the second-order Theil index. Finally,the spatial β convergence model was utilized to assess convergence patterns across the 269 cities, eastern, central and western regions, and 23 provinces (autonomous regions) in China. The research discovered that: ① Throughout the study period,the mean values of 269 cities fluctuated between 0.41 and 0.54,indicating relative stability and significant potential for enhancement according to theoretical projections. The average value of UGDE across the 269 cities,as well as in the eastern,central,and western regions,followed a U shaped temporal pattern,with a decline followed by an increase. Spatially,a high value in the east and a low value in the center were observed,with the region of Shanxi-Hebei-Shandong-Henan being a low-value cluster, and the southeast coast area being a high-value cluster. ② The UGDE show polarization and gradient characteristics,with the internal absolute differences in efficiency values widening over time. These differences mainly come from the increasing inter-provincial disparity after 2013. ③ When considering spatial spillover effects,the 269 cities, eastern, central and western regions, and 23 provinces (autonomous regions) display significant (absolute/conditional) β convergence characteristics. The order of convergence rate is eastern > western > central region, and the convergence rate of 269 cities in China is between central and western region. Provinces demonstrate diverse levels of significant (absolute/conditional) spatial β convergence tendencies,yet there are marked differences in the convergence timeframes. Additionally,the study reveals that conditional β convergence occurs faster than absolute β convergence,suggesting that UGDE will theoretically first converge toward its own steady-state equilibrium level. These findings provide a clear data profile for understanding the spatial patterns and evolution characteristics of China's UGDE,serving as a valuable reference for reducing efficiency imbalances. Compared with similar studies,there are two possible marginal contributions in this paper: Firstly,this paper incorporates the setting of a global common production frontier into the super SBM model,thereby enhancing the comparability of UGDE across different years. Secondly,the research considered the scale effect and spatial spillover effect in the convergence of UGDE,and conducted detailed tests of spatial convergence at multiple levels,including the national,regional,and provincial levels,which to some extent compensates for the lack of precision in previous convergence tests.