Abstract: Objective In recent decades, China"s rapid urbanization has led to large amounts of energy consumption and carbon dioxide emissions, triggering significant urban heat island effects and air pollution, which seriously threaten the physical and mental health of residents. How to improve the urban habitat has become the focus of urban planning. Urban ventilation, as an important way to deliver fresh and cold air to built-up areas, can effectively improve the comfort level. Although many international scholars have researched the optimization of urban ventilation, the existing research in China mainly focuses on the measurement of the current ventilation environment in cities. It neglects the long-term dynamic changes of the urban ventilation environment in response to different planning objectives in the process of urban development. Methods This study proposed a research framework with general applicability to simulating future changes in ventilation environment within city-region systems for the first time. Firstly, this study defined five indicators (land surface temperature gradient, roughness length, forest canopy density, elevation variation coefficient, and slope) to evaluate the ventilation environment from three dimensions: surface temperature gradient, surface roughness, and surface undulation. Secondly, a prediction model was constructed based on the random forest algorithm. The multiple scenarios were input into the validated prediction model to simulate changes in the future ventilation environment. Finally, depending on the historical trends in ventilation potential and differences across multiple scenarios within the study area, the applications were proposed for spatial planning and management of metropolitan areas. Results Influenced by the urban development from 2000 to 2020, the ventilation environment of the Taiyuan metropolitan area varies with time, space, and planning objectives. It showed a decreasing trend and an increasing trend in local areas. Patches of better ventilation environments occurred at Jinyang Lake, Fen River, and Park. Several potential ventilation corridors were formed on the east and west sides of the built-up area. Under the scenario of natural development, farmland protection, and ecological priority, the ventilation environment of the study area shows a gradual improvement trend from 2030 to 2050. Specifically, the ventilation environments in different regions show differences in spatial distribution. Under the natural development scenario, the extremely good and good ventilation areas show a trend of significant increase. With the continuous protection of basic farmland, the encroachment of urban expansion on farmland is effectively limited. Under the ecological priority scenario, the increase in good ventilated areas was concentrated in suburban forests and grasslands. Early fragmented ventilation areas gradually expanded and formed ventilation corridors around the built-up area. Conclusion In this study, a dynamic simulation model of the urban ventilation environment was constructed and the spatial distribution of the ventilation environment under different scenarios was mapped. To cope with a series of urban problems caused by urban sprawl and densification, the trends of the ventilation environment in the Taiyuan metropolitan area are predicted and identified, and in this way, a series of targeted optimization strategies are proposed to provide cooler and more comfortable environments for urban residents.