Fire protection of closed car parks, where it is possible to have a massive stay of people, is the most important and most crucial task when designing such facilities. The efficiency and effectiveness of smoke ventilation systems is achieved through compliance with design rules and numerical simulation of air distribution. The most reliable method is an experimental check of the efficiency of smoke ventilation. It is not possible to reproduce a real fire in the conditions of an object without damaging the structures of a building or a facility; therefore, tests are carried out with reduced, safe fire parameters. Based on the similarity theory, the article discusses the methodology for scaling test fire parameters and smoke ventilation parameters, which allows obtaining results relevant to the design fire parameters. In this case, the positive test results with lower parameters of the test fire confirm the design limits of the smoke distribution and the possibility of evacuating people from the car park in case of a fire having design parameters. An alternative to these scaling methods can only be the prediction of design test parameters and conditions based on the same calculation method, as well as the case when the prediction of the test results is confirmed by observations.