Given that measurements made under steady-state laboratory conditions cannot fully reflect the actual performance of air conditioners, the dynamic measurement of room air conditioner performance based on virtual buildings has become a research hotspot. This study determines the key parameters for the dynamic measurement of room air conditioner performance via simulation and model analysis methods, focusing on the heat load of the building, the zero-load point, the heat-to-cold ratio, the sensible-heat ratio, the selection coefficient, and the room's effective heat capacity and moisture capacity. The feasibility of this dynamic performance-measurement method is experimentally verified. The results show that for air conditioners applicable to average buildings in China, the selection coefficient for the cooling season was 1.33. For rooms equipped with an air conditioner with 2.4 kW nominal capacity (12-20 m²), the effective heat capacity was 240-300 kJ/K, and the moisture capacity was 50-100 kg. Based on these parameters, a dynamic performance test of a 3.5 kW air conditioner was conducted in the laboratory. These parameters resulted in a dynamic energy efficiency 31.8% lower than the steady-state energy efficiency, thus more accurately reflecting the actual performance of the air conditioner.