We apply a global sensitivity analysis of magnetic resonance sounding (MRS) data to the hydraulic parameters of a hydrological model at a catchment scale. The methodology used provides a quantitative estimate of the proportion to which the explored parameters influence the simulated MRS response. The sensitivity analysis is performed through a variance-based approach that provides indicators to infer how the MRS sensitivity to the different parameters might vary in space and time. We also apply the sensitivity analysis to water table depth estimates. We show that MRS measurements complement fruitfully classical water table depth acquisitions. Indeed, MRS is sensitive to the model water content at saturation while this parameter does not influence significantly the water table depth estimates.