Abstract The performance of heating and cooling systems are affected by their dynamic operating conditions. As a consequence, a sound evaluation of system performance should be done through a dynamic test procedure. However, the complexity, the cost, and the time for such type of experimentation are obstacles to the diffusion of this test method. A new whole system test procedure was developed to reduce its potential failure. The test method considers the emulation of system components that cannot be installed in the laboratory, while the selection of short-yet-significant test sequence is performed classifying the days of the year with clustering analysis. The procedure was applied to a solar-assisted heat pump system in four different European climates. The seasonal performance figures are extrapolated from the test results and compared with the numerical simulations of the entire heating and cooling system. In all test cases the seasonal performance factors are lower than the simulated ones by approximately 10%. The results obtained with dynamic tests confirm the necessity of their implementation, since they provide useful information to improve the system layout and control. In this way, the manufacturers can introduce more efficient products into the market.