Mindfulness-based cognitive therapy (MBCT) stands out as a promising augmentation psychological therapy for patients with obsessive-compulsive disorder (OCD). To identify potential predictive and response biomarkers, this study examines the relationship between clinical domains and resting-state network connectivity in OCD patients undergoing a 3-month MBCT programme. Twelve OCD patients underwent two resting-state functional magnetic resonance imaging sessions at baseline and after the MBCT programme. We assessed four clinical domains: positive affect, negative affect, anxiety sensitivity, and rumination. Independent component analysis characterised resting-state networks (RSNs), and multiple regression analyses evaluated brain-clinical associations. At baseline, distinct network connectivity patterns were found for each clinical domain: parietal-subcortical, lateral prefrontal, medial prefrontal, and frontal-occipital. Predictive and response biomarkers revealed significant brain-clinical associations within two main RSNs: the ventral default mode network (vDMN) and the frontostriatal network (FSN). Key brain nodes —the precuneus and the frontopolar cortex— were identified within these networks. MBCT may modulate vDMN and FSN connectivity in OCD patients, possibly reducing symptoms across clinical domains. Each clinical domain had a unique baseline brain connectivity pattern, suggesting potential symptom-based biomarkers. Using these RSNs as predictors could enable personalised treatments and the identification of patients who would benefit most from MBCT.