Theory of Mind (ToM) is the ability to infer other people’s mental states such as desires, intentions and beliefs. It constitutes a crucial prerequisite for many forms of social interaction and is impaired in various psychiatric and neurological diseases. ToM can be differentiated into affective (i.e. empathising the feelings of another person) and cognitive (i.e. inferring the mental state of the counterpart) subcomponents. While affective ToM abilities mainly seem to be mediated by the ventromedial prefrontal cortex, current research suggests that cognitive ToM abilities are associated with more lateral, in particular, dorsolateral prefontal regions. Recently, the basal ganglia (BG) are also ascribed to ToM. The multifaceted concept ToM inspires varied research investigating its neural underpinnings but few studies compare affective and cognitive ToM directly using functional imaging approaches of the brain. In order to specify the neural correlates of affective and cognitive ToM as well as the involvement of the BG, an adapted fMRI-version of the Yoni paradigm by Shamay-Tsoory and colleagues was applied. It contained an affective (aff) and cognitive (kog) ToM condition as well as a control condition (phy) and consisted of 60 highly comparable visual stimuli (20 in each condition). The items were shown for 6 sec and intermitted by a jittered fixation cross (from 3.5 to 4.3 sec, mean: 3.9 sec). Participants were asked to infer either the cognitive or affective mental state of the main character named Yoni whereas the control condition did not require any mental state inferences. 30 healthy right-handed participants (15 women, 15 men, mean age = 25.3 +/- 2.5 years) underwent event-related fMRI scanning (1.5 T, TE = 50 msec, TR = 3000 msec), neuropsycholo¬gi¬cal testing and filled in several questionnaires regarding general health and personality traits. In addition, three ToM tests were applied. Statistical analysis was conducted using SPM 5 and 8 for the functional imaging data and SPSS version 18 for the behavioural data. In both the contrasts, aff over phy and kog over phy, activation was found in regions previously ascribed to the ToM network, namely parts of the temporal lobe including the superior temporal sulcus, the supplementary motor area and parietal structures of the right hemisphere. The contrast aff over phy yielded additional activation in the orbitofrontal cortex on the right and the cingulate cortex, the precentral and inferior frontal gyrus on the left. Interestingly the right BG were recruited as well. Additional clusters of this contrast were located within the right parietal cortex including the precuneus and in the left cerebellum. The direct contrast aff over kog showed activation in the temporoparietal junction and the cingulate cortex on the right as well as in the left supplementary motor area. The reverse contrast kog over aff however did not yield any significant clusters. The contrast aff over phy revealed activation in the caudate nucleus and the pallidum of the right hemisphere. β-values derived from the anatomically defined region of interest of the BG correlated with the subscale perspective taking from the Interpersonal Reactivity Index. In spite of detailed neuropsychological testing and the application of several questionnaires only sporadic correlations between measures of ToM abilities and the collected behavioral data were found. The results of this study support the hypothesis that the neural correlates of affective and cognitive ToM subcomponents partly share neural correlates but can also be differentiated on an anatomical level. Affective ToM seems to recruit more regions than cognitive ToM and - consistent with previous research - more medial parts of the frontal cortex. Regions previously described as representing the ToM network were implicated in the present study. Furthermore, the BG seem to be involved in affective ToM possibly providing a motor component of simulation processes. The scarcity of correlations found between ToM and other cognitive measures or personality traits supports the hypothesis that ToM is a largely independent domain of social cognition. The investigation of the subcomponents of ToM helps to achieve a more detailed understanding of this complex construct that seems to comprise very different abilities.