Magnetic resonance imaging (MRI) has tremendous potential for revealing transport processes in engineered and geologic systems. Here, we utilize MRI to image nanoparticle (NP) transport through a saturated coarse-grained system. Commercially available paramagnetically tagged NPs are used; the paramagnetic tag making the NP visible to MRI. NP transport was imaged as NPs migrated through packed columns of quartz and dolomite gravel. Changes in T2-weighted image intensity were calibrated to provide fully quantitative maps of NP concentration at regular time intervals (T 2 being the spin–spin relaxation time of 1H nuclei). Transport of nanoparticles was significantly retarded in dolomite compared to quartz due to electrostatic attraction between nanoparticle and dolomite surfaces. NP concentration profiles were evaluated with the CXTFIT computer package to estimate nanoparticle transport parameters at multiple points along the length of the column. This provided temporally resolved parameters that standard breakthrough curve analysis cannot provide. Particle–surface interaction energy profiles were described through Derjaguin-Landau-Verwey-Overbeek (DLVO) theory. While dispersion coefficients and fast deposition rate constant (k fast) were found to increase with distance, deposition rate constant (k) and collision efficiency (α) were found to decrease with distance. These length-dependant variations have significant scaling-up implications for transport models used to predict NP transport in natural and engineered coarse-grained systems, such as sustainable urban drainage systems and river beds.