The highly heterogeneous hydration dynamics of protein-water interfaces is considered important for protein stability and dynamics, protein folding, enzymatic activity, and even drug design. The nuclear Overhauser effect (NOE) between protein and water protons is the only experimental observable which, in principle, can provide a map of locally resolved hydration dynamics. However, its utility was questioned in various theoretical studies that emphasized the contributions of long-range NOE interactions. We show by a detailed analysis based on molecular dynamics simulations that, contrary to recent claims, the protein-water NOE is an excellent observable to map local hydration dynamics at the protein surface.