This study addresses the controversy over how motor maps are organized during action simulation by examining whether action simulation states, that is, motor imagery and action observation, run on either effector-specific and/or action-specific motor maps. Subjects had to observe or imagine three types of movements effected by the right hand or the right foot with different action goals. The functional magnetic resonance imaging results showed an action-specific organization within premotor and posterior parietal areas of both hemispheres during action simulation, especially during action observation. There were also less pronounced effector-specific activation sites during both simulation processes. It is concluded that the premotor and parietal areas contain multiple motor maps rather than a single, continuous map of the body. The forms of simulation (observation, imagery), the task contexts (movements related to an object, with usual/unusual effector), and the underlying reason for performing the simulation (rate your subjective success afterwards) lead to the specific use of different representational motor maps within both regions. In our experimental setting, action-specific maps are dominant especially, during action observation, whereas effector-specific maps are recruited to only a lesser degree.