In many ATM studies experiments are performed to determine the capacity. This paper looks at the effect of airspace design on the capacity. Using an algebraic approach a relation is derived between the design parameters of a layered airspace design and the capacity of the airspace. The validity of the assumptions which are used in this derivation are tested experimentally. This airspace lay-out proved to be the airspace design which had the highest capacity for the unstructured, extremely high traffic demand used in an earlier experimental study. The result is both a method to relate an airspace design to the capacity as well as a relation which shows the effect on the airspace capacity for an airspace design where different levels or layers are defined each with their own segment of heading angles.