This contribution presents an approach to use fuzzy logic and fuzzy rule systems for enterprise application evaluations (enterprise architecture dependent application evaluations). Such evaluations regard all relations to enterprise architecture artifacts in order to evaluate an application. There may be a large number of criteria to be considered and some of them have to be assessed by humans, which may introduce vagueness to the results. We present a method to assign the results of criteria to linguistic terms that have semantics like “good” or “poor”. Since it is often relevant to have exactly one characteristic to compare applications to each other, a fuzzy rule system approach is presented that aggregates those linguistic terms for each criterion of an application to an overall indicator of an application’s quality.