Background Major Depressive Disorder (MDD) is a common psychiatric illness. Neuroimaging studies have indicated the involvement functional networks in the disorder. Resting-state functional magnetic resonance imaging (rs-fMRI) provides a non-invasive, task-free way of studying such distributed functional dynamics of the brain. Previous studies have found significant alteration in rs-fMRI functional connectivity in MDD patients. Hyperactivity of default mode network (DMN) has been reported to be associated with rumination of negative emotional stimuli, and abnormal connectivity between DMN and task-positive networks was associated with cognitive impairment within MDD, however other studies do not report such effects. Previous rs-fMRI studies were however often conducted on small sample sizes which may account for inconsistency in the literature, and few studies were conducted on the association between genetic risk of MDD and rs-fMRI functional connectivity. Therefore, the present study used imaging data from a large-population based sample from UK Biobank to test the association of brain functional connectivity with MDD status and its genetic risk. Methods We used IDPs (Imaging-Derived Phenotypes) from UK Biobank imaging project. The imaging data was collected on a single scanner and preprocessed using FSL packages by UK Biobank. Data was parcelled into 21 non-noise functional networks, and the L2-regularised partial correlation was estimated between each pair of networks. These estimated connections were then used for further analysis in the current study. MDD status was derived based on self-reported depressive symptoms and history of hospital admission data (Case: N=256, Control: N=488,). Polygenic risk of MDD (MDD-pgrs) was calculated based on the summary statistics from the PGC (N=914). Regression models were applied to test the MDD case-control difference and the associations of MDD-pgrs with rs-fMRI functional connectivity. Age, age2, sex and assessment centre were controlled for in the model of case-control differences, and genotype batch and the first 15 principal component of genetic variance were additionally controlled for MDD-pgrs. FDR correction was applied. All effect sizes reported were standardised. Results MDD cases showed higher positive connectivity between the precuneus, the bilateral superior posterior cingulate cortex and temporal-parietal junction (β=0.292, pcorr=0.047); all components of DMN. Higher MDD-pgrs was associated with less connectivity between posterior motor area and posterior parietal gyrus (PPG) (β=-0.123, pcor=0.045), less negative connectivity between precuneus and PPG (β=0.122, pcor=0.045) and higher negative connection of right fronto-temporal network (FTN) and left fronto-parietal network (FPN) (β=-0.124, pcor=0.045). Discussion In the present study, MDD cases exhibited significantly higher connection between two DMN components, and a higher negative connection between FTN and FPN. Conversely, higher genetic risk for MDD was associated with decreased connectivity between DMN regions and task-positive networks. These results were highly consistent with previous studies which found that altered functioning of DMN was related with MDD status. The present study also found functional connections that involve with DMN, FTN and FPN associated with MDD-pgrs.