This paper focuses on flexibility in the area of flexible automatic assembly systems. Far-reaching research has been done in effort to develop flexible automatic assembly systems. A dilemma that appears in understanding the focus for these research projects is the immense amounts of flexibility concepts used. This paper proposes a structure for the concepts based on frequency and horizon for different events in an assembly system. The results are presented in a proposed “flexibility map”. Based on the “frequency and horizon approach” and the definition of flexibility as “the penalty of change” [1], an argumentation concerning what kind of flexibility is most valuable in an assembly system can be discussed. The paper also points out the demands for a flexibility that allows for de-coupling of high and low frequency events in a system. A clear distinction is made between dynamic flexibility and static flexibility. It is a well-known fact that the increase of the dynamic flexibility in an assembly system will decrease the static flexibility [2]. The paper ends with a proposal for decreasing the negative correlation between dynamic- and static flexibility. An objective for further research concerning whether a certain process within an assembly system affects certain flexibility is also presented.