More than three-fourth of chemotherapy treated cancer survivors have cognitive impairment, including memory loss, inability to think, lasting several years after completion of therapy^1,2. There is also evidence of higher risk of dementia among chemotherapy treated cancer survivors³. Recently, functional neuroimaging has played an important role in unraveling the complex underpinnings of this cognitive decline associated with chemotherapy, sometimes labelled as ‘chemobrain’⁴. Default mode network (DMN) is a functional brain network that deactivates during high cognitive demand tasks and activates during “rest”⁵. The regions in the DMN have been found to be critical hubs that connect several networks of the brain. DMN has been implicated to change in several brain disorders. Further study of DMN in chemotherapy-induced cognitive decline could aid in development and improvement of therapy. DMN has been shown to be specifically vulnerable to chemotherapy^6,7. DMN connectivity has been shown to be disrupted following chemotherapy⁸ and DMN connectivity features were found to contribute significantly to classify chemotherapy patients from non-chemotherapy patients⁷. Changes in DMN post-chemotherapy with baseline measurement prior to chemotherapy has been suggested, but the authors are unaware of any report on longitudinal changes in resting state functional connectivity of the DMN. In this study we study changes in DMN connectivity after chemotherapy within the same patients.After obtaining informed consent, resting state fMRI scans were acquired from six pre-menopause breast cancer patients prior to and after (3 months to 1 year after initial scan) chemotherapy treatment. Resting state fMRI was performed on a 3T Signa HDxt MR scanner (GE Healthcare, Waukesha, WI) using an eight-channel phased array brain coil for 7 minutes with TR/TE of 2000/30 ms, whole brain coverage, voxel dimensions of 1.875 x 1.875 x 5.5 mm. Data from one subject was noisy and hence excluded from analysis. In all but two sessions, two rs-fMRI scans were acquired equivalent to 14 minutes of data. T1-weighted image was also obtained for spatial normalization purposes. The rs-fMRI data acquired were motion corrected, rigidly registered to the T1-weighted image, non-rigidly registered to MNI atlas, nuisance corrected using aCompCor⁹ including removal of global signal, spatially smoothed with a Gaussian of FWHM 4 mm and temporally filtered to pass between 0.01 and 0.1 Hz using in-house custom built software. Functional connectivity map of the default mode network was extracted using correlation with a seed region of 6 mm radius in the posterior cingulate cortex¹⁰. The difference in connectivity estimates (post-chemotherapy minus pre-chemotherapy) was computed and tested for statistically significant (p<0.05 after multiple comparison correction) change. Typical DMN was observed (as shown in Figure 1) in pre-chemotherapy patients with PCC connections to bilateral inferior parietal lobules, ACC, medial temporal lobes and posterior cerebellum. Group average of DMN is shown in Figure 2 with typical regions with no visible difference in connectivity. Within each subject the difference between post and pre-chemotherapy showed significantly increased connectivity of the PCC to bilateral hippocampus and the thalamus (Red-yellow regions in Figure 2). Significant connectivity decrease in correlation to task positive networks were observed (blue regions in Figure 3). This decrease in correlation is actually an increase in anti-correlation post-chemotherapy. Memory related problems are the most frequently reported cognitive impairment following chemotherapy¹. The DMN has been previously shown to be specifically involved in chemotherapy related changes in cognition. The DMN is also functionally connected to the hippocampus¹⁰, a region strongly associated with memory. Increased connectivity with bilateral hippocampus may be associated with the memory impairments due to chemotherapy. The dorsal attention network which includes the intra-parietal sulcus and the DMN are typically anti-correlated with each other and is claimed to be competitive networks¹¹. The degree of anti-correlation has been shown to be related to behavioral characteristics¹² including autistic traits and vigilance. Here we observe an increase in the anti-correlation, which probably reflects an increased level of vigilance. The increased anti-correlation could also be a compensatory effect to the cognitive decline. Future directions include validating this in larger data and associating the connectivity differences to psychometric scores related to specific cognitive functions.