<b><i>Background:</i></b> The estimated glomerular filtration rate (eGFR) is frequently used to monitor progression of kidney disease. Multiple values have to be obtained, sometimes over years to determine the rate of decline in kidney function. Recent data suggest that functional MRI (fMRI) methods may be able to predict loss of eGFR. In a prior study, <i>baseline</i> data with multi-parametric MRI in individuals with diabetes and moderate CKD was reported. This report extends our prior observations in order to evaluate the temporal variability of the fMRI measurements over 36 months and their association with annual change in eGFR. <b><i>Methods:</i></b> Twenty-four subjects with moderate CKD completed 3 sets of MRI scans over a 36-month period. Blood oxygenation level-dependent (BOLD), arterial spin labeling perfusion, and diffusion MRI images were acquired using a 3 T scanner. Coefficients of variation was used to evaluate variability between subjects at each time point and temporal variability within each subject. We have conducted mixed effects models to examine the trajectory change in GFR over time using time and MRI variables as fixed effects and baseline intercept as random effect. Associations of MRI image markers with annual change in eGFR were evaluated. <b><i>Results:</i></b> Multi-parametric functional renal MRI techniques in individuals with moderate CKD showed higher temporal variability in R2* of medulla compared to healthy individuals. This was consistent with the significant lower R2* in medulla observed at 36 months compared to baseline values. The results of linear mixed model showing that R2*_Medulla was the only predictor associated with change in eGFR over time. Furthermore, a significant association of medullary R2* with annual loss of eGFR was observed at all the 3 time points. <b><i>Conclusions:</i></b> The lower R2* values and the higher temporal variability in the renal medulla over time suggest the ability to monitor progressive CKD. These were confirmed by the fact that reduced medullary R2* was associated with higher annual loss in eGFR. These data collectively emphasize the need for inclusion of medulla in the analysis of renal BOLD MRI studies.