This report focuses on the S-Box used in CMEA . The purpose is to find the generation process of this S-Box. Such knowledge is important since a weak design process usually results in vulnerability against attacks. Many of them have already been published against CMEA. The attackers used the statistical bias in the S-Box to develop their attack. We want to know what caused such a statistical bias, so that this method may be avoided later on. In order to find a structure in this S-Box, we first recall the high level structure that has already been found before. We then look at several properties, such as the linear and differential properties and the hardware related properties, and finally, the relations with other cryptographic algorithms. Our results show that the TU structure is likely just a consequence of the S-Box being generated from thirty-two 4×4 permutations being concatenated. The properties of the tables components involved in the structure are all consistent with those of pseudo-random permutations. Hence, the S-Box might just be constructed using thirty-two pseudo- randomly generated4×4 permutations.