Most of us can easily conjure up fictitious scenarios in front of our mind’s eye with great clarity. Neurally, these types of mental events rely on a neural network involving the ventromedial prefrontal cortex (vmPFC), hippocampus and visuoperceptual cortex. Whereas these brain regions have been associated with mental event construction earlier, the precise contributions of each node are hotly debated. For example, the vmPFC is usually associated with abstract, schema-like mental events, however we proposed earlier that the vmPFC may initiate and elaborate vivid, detail-rich mental events. Therefore, in this functional magnetic resonance imaging (fMRI) study, we pitched three closely matched cognitive tasks against one another: scene construction, single object construction, and an abstract definition. For each of three words, (e.g., hope, secret, order), participants learned three mental events: 1) a scene event, 2) a single object, and 3) an abstract semantic definition. During the scanning session, we repeatedly prompted participants with combinations of the cue words and learned conditions (i.e., Hope – scene, Secret - definition). We found greater activity for the scene condition in the medial frontal gyrus, medial temporal lobes and posterior cingulate gyrus. During the object condition, we found greater activity in the inferior frontal gyrus, and the occipital cortex, and during the abstract definition condition, we obtained high activation in superior temporal gyrus, and parietal gyrus. These results shed new light on the precise contributions of various brain structures during the construction of the movies of our minds.