It is typically assumed that impaired structure or structural connectivity will have corresponding influence on functional connectivity and neural activation. Indeed many clinical studies of disease severity use fMRI metrics as biomarkers [1]. We hypothesized that structure and functional responses do not demonstrate the same pattern of impairment across time. Using the cuprizone mouse model of reversible demyelination [2], we tracked the progression of neural activation and functional connectivity, and compared the longitudinal changes against that of healthy mice and also the extent of cuprizone demyelination using histology.Experimental design: 8-week-old C57BL/6 male mice (n=8) were fed 0.2% (w/w) cuprizone mixed in milled chow for 6 weeks, then switched to normal chow for another 6 weeks. They were imaged at four time points: before starting diet (week 0), 3rd , 6th and 12th week from start of cuprizone diet. We used a 9.4T scanner (Agilent Technologies, USA), under a protocol optimized for mouse resting-state fMRI (0.1-0.6mg/kg/h medetomidine i.p. [3]). fMRI of both resting state and forepaw stimulation (6Hz, 0.5mA) was performed. A control group of healthy C57BL/6 mice (n=8) were scanned with the same protocol at the same time points. A 3rd group of C57BL/6 mice (n=9) went through the same cuprizone regimen and were sacrificed at 9 different time points to assess myelination using Luxol fast blue staining. MRI: BOLD EPI: TR/TE=2000/15ms, α=90°, slices=15, voxel=0.31x0.31x0.5mm3, volumes (resting state / forepaw stimulation) = 300/150. FSE: TR/TE=2500/40ms, echo-train=8, average=2, voxel=0.08x0.08x0.15mm3. Analyses: Postprocessing: Data were motion corrected with SPM. Ventricle and muscle signals were regressed out to avoid physiological noise. All data were coregistered onto an anatomical template using FSL FLIRT and FNIRT, followed by spatial smoothing with Gaussian kernel of 2 pixels FWHM. Resting state fMRI: Data were highpass-filtered at 0.01 Hz. Whole-brain connectivity maps were generated by seed-based correlation of averaged regions based on anatomical template and transformed using Fischer’s z-transformation. One sample t-tests were then done to determine significant connectivity on a group level. The connectivity scores (t-values) from M1, S1, S2 and V1 were averaged (cortical regions), while connectivity scores from the CPu, GP, thalamus, and hippocampus formed another average (subcortical regions). Forepaw stimulation fMRI: GLM was performed to get BOLD activation maps with threshold of p<0.01. Number of activated voxels and BOLD response amplitudes were calculated. Statistics: Mixed model ANOVAs and paired t-tests were done to determine significant differences.Demyelination occurred with the introduction of the cuprizone diet and progressed steadily till Week 6, after which cuprizone was withdrawn and remyelination occurred as shown in Week 12 (see Fig. 1 for the histology and T2-weighted images showing the demyelination progression). A different pattern was observed for interhemispheric connectivity (Fig. 2): An oscillating pattern of connectivity can be seen for the cuprizone group, with the weakest connectivity at Week 3, increasing at Week 6 and declining again at Week 12. This was in all regions observed. There was a significant interaction between group and time points (p<0.05) with significant difference between Weeks 0-3 (cortical), Weeks 3-6 (subcortical), Weeks 3-12 (subcortical). The control data, which was averaged from all the regions showed no significant difference across time, except for Week 0. This was attributed to a global 30% drop in tSNR during those control scans. Yet a third pattern was observed for neural activation (Fig. 3). Both the activated voxel count and BOLD response amplitude show significantly elevated responses (p<0.05) at Week 3 and Week 6 before returning to baseline in Week 12. The control data showed no significant changes across time.We show that the trends of functional responses do not necessarily follow the trend of structural injury and recovery. In the typical cuprizone regime, demyelination occurs progressively until the drug is withdrawn at 6 weeks. However we observe that functional connectivity oscillates from initial impairment to subsequent increase and settles down again with remyelination. On the other hand, neural activation is elevated during the cuprizone regime and returns to baseline upon remyelination. Possible reasons for the heightened activation responses include inflammation and disrupted neurovascular coupling. The mild swing back of functional connectivity at 6 weeks, even when demyelination is most severe, is reminiscent of the increases in functional connectivity seen in multiple sclerosis patients with white matter damage [4, 5]. In the latter study, stronger functional connectivity also correlated with worse cognitive scores. The contrasting patterns observed in this study highlight the complex relationship of structure and function, and the need to exercise caution when using functional imaging biomarkers in assessing disease progression.