Underground space is an important asset for urban development, but its development comes with inherent risks. Therefore, in order to evaluate the risk of underground space development, we utilized 3D modeling and grey relational analysis (GRA) for decision-making, and evaluated underground space development through attribute modeling. First, the underground space was categorized into 4 layers according to depth and 15 important influencing factors were identified. Then, the weights of these factors (soft soil thickness, liquefaction index of saturated sand, etc.) were determined using analytic hierarchy process (AHP), criteria importance through intercriteria correlation (CRITIC), and GRA. Finally, the study area was divided into 1,197,778 evaluation units, each measuring 30 m × 30 m. The suitability of each unit was evaluated using the Multiple Criteria Decision Analysis (MCDA) combined with the GRA method. The experimental results show a commendable performance compared to the combined “AHP + multilevel exponential superposition” approach. In addition, the computing efficiency is reduced by about 30%, the stability is improved by at least 20%, and the reliability is improved by about 10%. The study enhances the decision-making integration of GRA, MCDA and underground space, improves the efficiency and accuracy of underground space risk evaluation, and provides a valuable basis for the field related to underground space development. Code is available at https://github.com/NX-first/underground-space-engineering-development.git.