Test and evaluation of wireless communication performance with cars can be done in many ways. One approach, for both vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) tests, is to use a multipath propagation simulator to achieve realistic signal environments for the different communication systems under test. Multipath propagation simulators have previously been shown to be useful for testing mobile handsets. Specific issues related to tests with cars are the large size of the test object and the disturbances it can cause. In this paper we experimentally study how different design parameters of the multipath propagation simulator affect the received signals. The analysis of the setup as well as the data is performed using the so-called 'Design of Experiments' method.