Fuzzy Cognitive Maps are very useful tools primarily in decision making and management tasks. They represent the main factors, variables of a complex system and the internal causal relationships among them in a straightforward way. Simulations can be started with an initial state, and the future states of the system under investigation can be predicted. This way, what-if questions can be answered. If the model of a system is created by experts they are often tempted to include too many components, because they are not sure in the importance of them. An oversized model is excruciating to use in practice, however. Model reduction methods help to decrease model size but unavoidably cause information loss as well. This effect does not cause a problem in practical decision making applications if the model suggests the same decisions. This chapter covers three FCM model reduction methods, their theoretical background and behavioral properties.