Emotion regulation is a key concept in psychopathology and psychotherapy theoretical models (Aldao et al., 2012; Greemberg & Pascual-Leone, 2006). It has been taken into consideration as an important issue in the debates about the possibilities and limits of neuroimaging techniques application in studying complex clinical phenomena (Westen & Gabbard, 2002; Kandel, 2007). In neuroscience, emotion regulation has been described in terms of cognitive control of emotional responses, in line with the dual-process models based on the difference between controlled and automatic processes (Barrett et al, 2004). In the first chapter, a meta-analysis of fMRI studies on neural correlates of emotion regulation is described with the aim to summarize the data from the literature, and, in the second place, to compare adaptive and non-adaptive forms of emotion regulation. The results confirm the previous literature that has shown the increase of prefrontal areas activation and the decrease of limbic areas activation in carrying out voluntary strategies of emotion regulation (Ochsner et al., 2008; Diekhof et al., 2011). However, the comparison between adaptive and non-adaptive emotion regulation strategies did not show relevant differences. These results are discussed taking into account the studies’ methodological limits and the intervention of processes that go beyond the categorization of the dual-process models. Dual-process model of emotion regulation has been applied in describing psychotherapy effects. In this case, it has been hypothesized a change in emotion regulation brain areas as effect of psychotherapy intervention (De Rubeis et al., 2008). With the aim to verify this hypothesis, in the second chapter the results of a second meta-analysis of longitudinal studies on neural effects of psychotherapy is exposed. In line with the results, only a partial confirmation to the dual-process model was provided, whereas it was shown the intervention of semantic attribution processes, dependent on temporal areas (Binder et al., 2009), and self-related information processes, depending on the cortical midline structures (Northoff et al., 2006). In order to enrich the dual process models of emotion regulation, a neuroimaging study, described in the third chapter, had the aim to explain the implication of attentional processes not attributable to cognitive control. With this aim, the neural correlates of proactive interference (the interference of responses that were previously correct but no longer are) have been investigated. Proactive interference resolution require the control of the internal informational environment, as opposed to resisting the effects of salient or prepotent stimuli of external origin in the elaboration of response, and it is related with semantic categories of the experimental stimuli, which in this experiment included emotional or neutral categories. The result of this study have shown that the emotional content of the experimental stimuli interacts with the increase of proactive interference affecting the modulation of cerebral areas part of the semantic system, (Binder et al., 2001), of the attentional ventral network, involved in forms of attention dependent on the behavioural relevance of stimuli, more than on voluntary attention (Corbetta et al., 2008), and the default system associated with internal directed attention (Raichle et al., 2001). In the fourth chapter, an analysis of individual differences is described, which showed the association between individual differences in spontaneous avoidance and the modulation of temporal semantic areas. Finally, at the conclusion of the present dissertation, some clinical and conceptual remarks related to the empirical data are discussed.