The Chair of Production Systems conducts many reliability and fatigue tests of SMA actuators. A properly functioning test rig is essential for these often tedious investigations. The authors found deviations in the experimental results over high amount of experiments e.g. in the case of a batch change of the test material or in the case of exchanging wear parts. It is not clear whether this change was actually measured or is caused by a deviation of the measuring equipment. The influences of different load parameters on structural and functional fatigue are only partially quantified, which makes it difficult to assess the general plausibility of the results. In this paper the authors propose an experimental design by means of design of experiments (DoE), which determines the qualitative influence of the independent variables (IV) mechanical load, activation time, voltage, current and energy input on the dependent variable (DV) fatigue. The results of this initially performed experimental design serve as a reference point during the operation of the test rig in the event that experimental results do not appear plausible. In this case, the reference experimental design can be repeated. Since the influence of each IV is represented separately by the DoE, not only a fundamental deviation can be determined, but also the IV which deviates from the reference can be determined directly.