ObjectiveAttention Training Technique (ATT) is a key therapeutic tool in metacognitive therapy. There are numerous studies on the behavioral effects of ATT, however the neural mechanisms at work in the training are yet to be uncovered. To date there have been no controlled fMRI studies of ATT. MethodWe conducted a randomized double-blind controlled study of two groups with varying levels of cognitive-attentional syndrome (CAS). Groups with high (n = 43) and low (n = 46) levels of CAS underwent a single session of ATT or a control condition (CON) in an MRI scanner. Participants underwent resting state functional MRI (rsfMRI) sessions and rumination induction sessions both pre- and post-intervention Functional connectivity analyses and inter-subject correlations analyses were computed. We also collected data on emotion and attention functioning pre- and post-intervention. ResultsWe did not observe any behavioral effects of ATT. However, direct comparison between ATT and CON sessions revealed greater inter-subject correlations in almost all hubs belonging to the studied functional networks. Moreover, subjects who received ATT showed diminished connectivity in the fronto-parietal network during ruminations and diminished connectivity of the precuneus with lateral occipital cortices and the intraparietal sulcus in abstract thinking and rsfMRI, respectively. Furthermore, some of the observed effects in functional connectivity and inter-subject correlations were specific to different levels of CAS. ConclusionsOur results may support a proposed neural mechanism for ATT: disengagement of attention from CAS-type processing in either low- or high-CAS individuals. It is also possible that some neural effects of ATT are specific to individuals with different levels of CAS.