Reasonably assessing green transformation efficiency (GTE) and clarifying influencing factors can provide theoretical support for their sustainable development. This study utilizes the undesired superefficient Slack Based Measure model (SE-SBM), window analysis, kernel density estimation analysis, and Tobit model to assess the GTE of 39 resource-based cities (RBCs) in the Yellow River Basin (YRB), explore their dynamic evolution trends and influencing factors, and attempt to compensate for a lack of clarity in green transformation constraint factors. Following findings: (1) YRB’s GTE showed a “V-shaped” upward trend. There were differences between upper, middle, and lower cities: upper cities are higher. (2) GTE is evolving to a higher level, and the inter-regional equilibrium level has improved. The kernel density curve in the upper, middle, and lower reaches has its own regional characteristics and time period features. (3) Industrial structure upgrading, economic development level, and green technology innovation level are positive effects, while the opening-up level is negative. According to the findings, YRB’s RBCs should adjust measures to the current environment and urban conditions, promote digitization and intelligentization, and improve innovative economic growth, thus lifting the quality of green development. These findings also illustrate how the analysis framework mentioned in the study can enhance the understanding of urban green transformation and serve sustainable urban development.