A previously developed methodology for reducing the physical scale of test articles for fire performance testing of rail car floors was benchmarked using a series of large-scale tests and validated finite element simulations. The simulations showed that the experimental differences between the full-scale and reduced-scale test articles was due to testing discrepancies in the actual thermal and mechanical boundary conditions observed during the tests. While not directly validating the reduction methodology, this effort found no fundamental flaws in the methodology if the reduced-scale tests are properly executed.