Protein dynamics related to function can nowadays be structurally well characterized (i.e., instances obtained by high resolution structures), but they are still ill-defined energetically, and the energy landscapes are only accessible computationally. This is the case for glucose-galactose binding protein (GGBP), where the crystal structures of the apo and holo states provide structural information for the domain rearrangement upon ligand binding, while the time scale and the energetic determinants for such concerted dynamics have been so far elusive. Here, we use GGBP as a paradigm to define a functional conformational landscape, both structurally and energetically, by using an innovative combination of paramagnetic NMR experiments and MD simulations. Anisotropic NMR parameters induced by self-alignment of paramagnetic metal ions was used to characterize the ensemble of conformations adopted by the protein in solution while the rate of interconversion between conformations was elucidated by long molecular dynamics simulation on two states of GGBP, the closed-liganded (holo_cl) and open-unloaded (apo_op) states. Our results demonstrate that, in its apo state, the protein coexists between open-like (68%) and closed-like (32%) conformations, with an exchange rate around 25 ns. Despite such conformational heterogeneity, the presence of the ligand is the ultimate driving force to unbalance the equilibrium toward the holo_cl form, in a mechanism largely governed by a conformational selection mechanism.