The increased use of mechanical cooling systems in buildings due to factors like global warming and higher thermal loads has led to the recognition of natural ventilation (NV) as a sustainable alternative, offering lower operational costs and energy savings. However, NV’s effectiveness is limited in large spaces, such as sports complexes and indoor pools. This study presents a multiscale strategy to evaluate NV in such spaces, focusing on the Roman Baths Museum, with a volume of 11,000 m3, four pools, and 130 radiators. The aim of this study is to develop and validate a simplified CFD methodology that combines CFD simulations with in-situ experimental data to assess NV performance in large spaces. The methodology includes meso- and macro-scale models to examine thermal plume behavior and optimize NV design. A porous media model was used for simulating airflow and temperature distribution. Experimental validation was performed at the meso-scale, followed by macro-scale CFD simulations of the full-scale museum. The results showed relative errors below 10%, validating the model at both scales. The study found that natural ventilation is effective at lower elevations, but improvements in airflow mixing at higher levels are needed. This multiscale approach provides an efficient tool for optimizing NV in large buildings, offering valuable insights into future energy-efficient architectural designs.