Of all the nuclei in proteins, the nuclear magnetic resonance (NMR) chemical shifts of nitrogen are the theoretically least well understood. In this study, quantum chemical methods are used in combination with polarizable-continuum models in order to show that consideration of the effective electric field, including charge screening due to solvation, improves considerably the consistencies of statistical relationships between experimental and computed amide (15)N shifts between various sets of charged and uncharged oligopeptides and small organic molecules. A single conversion scheme between shielding parameters from first principles using density functional theory (DFT) and experimental shifts is derived that holds for all classes of compounds examined here. This relationship is then used to test the accuracy of such (15)N chemical shift predictions in the cyclic decapeptide antibiotic gramicidin S (GS). GS has previously been studied in great detail, both by NMR and X-ray crystallography. It adopts a well-defined backbone conformation, and hence, only a few discrete side chain conformational states need to be considered. Moreover, a charge-relay effect of the two cationic ornithine side chains to the protein backbone has been described earlier by NMR spectroscopy. Here, DFT-derived backbone amide nitrogen chemical shifts were calculated for multiple conformations of GS. Overall, the structural dynamics of GS is revisited in view of chemical shift behavior along with energetic considerations. Together, the study demonstrates proof of concept that (15)N chemical shift information is particularly useful in the analysis and validation of protein conformational states in a charged environment.