Experiments are the heart of the advancement of human knowledge and for an experiment to be informative it has to be properly designed. The objectives of many experiments are to study the e ects of several factors on a response, leading to Factorial Experiment. The size of a factorial experiment increases rapidly with the number of factors involved and the number of levels per factor. A major concern is to reduce the size of the experiment, while keeping it as informative as possible. This can be achieved by an important and widely used class of factorial experiments known as two-level factorial experiments where each factor is at two levels. Further reduction can be achieved by performing only a fraction of the experiment resulting in a Fractional Factorial Experiment. An important aspect of experimentation is the randomization mechanism. In many cases some factors are harder to randomize than others leading to Split-Plot Designs. In fact, there are many fractional factorial Split-plot designs (FFSP designs) that can be chosen for the same experiment. In this study, we focused on methods of selecting best designs using the maximum resolution criterion, the minimum aberration, and the minimum aberration secondary. In addition, we propose an additional criterion to select best design among the MA FFSP designs taking into account the cost of experiment.