Major Depressive Disorder (MDD) is a prevalent psychiatric disorder, characterized by high relapse risk. With every new episode, risk for relapse increases. This makes preventing relapse an important clinical target in limiting the personal and societal burden of MDD. Preventive Cognitive Therapy (PCT) is a protocolized psychological therapy which has shown to lower relapse risk. How PCT attains its effects needs further elucidation. Understanding the treatment mechanisms provides a window to identify critical target points to prevent depressive relapse. In this randomized controlled trial, 50 patients remitted from at least two depressive episodes in the past five years were randomized to eight sessions of PCT (n=25) or to a waiting list condition (n=25) in the context of the NEWPRIDE trial. Primary outcome measures were changes in brain activation during effortful emotion regulation and in biased processing, covering both negative and positive valence dimensions. All patients were assessed twice (baseline and three-month follow-up) for these outcome measures, as well as their diagnosis, symptomatology, cognitive and affective reactivity, and emotion regulation styles. Linear Mixed Models and Repeated Measures ANOVAs were conducted to objectify the immediate changes induced by the therapy in brain reactivity, and clinical and cognitive measures. Following PCT, patients showed decreased recruitment of dorsomedial prefrontal regions during upregulation of positive affect and stable recruitment of the pregenual anterior cingulate cortex during regulation of emotions over valences, compared to the waiting list. No effects on biased processing of emotional information were observed. Furthermore, PCT resulted in a lower increase of depressive symptomatology over three months as compared to the waiting list condition. Finally, PCT resulted in increased activation of positive thoughts following reading positive self-related scenarios, lower responsivity of negative affect to negative stimuli and increased successful application of cognitive reappraisal to modify affective states. These results suggest that PCT obtains its relapse preventing effects by targeting mechanisms that underpin regulation of mood. More specifically, changes in regulation of positive affect and content of positive cognitions may decrease negative mood and affect. This supports cross-valence compensatory models of cognitive therapy and suggests that strengthening and shifting cognition and affect to more positive content may guard against the activation of negative cognitions and affect in the face of daily hassles and life events.