Following economic growth in the past three decades, rapid urbanization has caused many pronounced issues, such as spatial scarcity and cultural discontinuity, in Chinese historical and cultural cities. In order to better deal with the diversification of underground space resources, data and information, this study introduces a random forest algorithm and proposes a multi-layer information superposition method. According to the characteristics of different information, starting from qualitative and quantitative aspects, we explore the effective performance of the rational development of underground space resources. Taking Yangzhou City, China, as an example, this paper evaluates the suitability and calculates the development volume of urban underground space. The development capacity, potential value, and comprehensive quality of underground space resources are explored in an attempt to demonstrate the practicality and scientificity of the evaluation method for achieving the developmental goals of urban space reconstruction and historic preservation. On this basis, an underground space scale forecast is carried out to provide decision support for relevant planners, managers, and construction personnel that is conducive to the orderly development of urban space, alleviation of increasing human–land conflicts, and coordination of the protection and development of underground space resources in historical and cultural cities, ultimately promoting sustainable development of cities.