We present a novel saturation transfer difference (STD) experiment where group selective (GS) saturation of amide protons in (15)N labeled hosts is achieved. It is demonstrated that a train of BIRD(d) pulses that inverts only protons attached to (15)N indeed results in saturation of the amide protons, while the background proton magnetization is much less affected. The undesired effect of partial saturation of the unlabeled protons can be completely cancelled out in difference spectra by switching the (15)N carrier between the on- and the off-resonance frequencies. As a result, clean and artifact-free STD spectra are obtained without the need of time-consuming optimization of experimental parameters and acquiring control spectra in the absence of the host. The use of the (15)N-GS STD experiment is demonstrated for the case of a glycopeptide antibiotic (dimeric eremomycin)-cell-wall analogue peptide (N-Ac-D-Ala) model system where the host and guest (1)H signals overlap. The application seems feasible for ligand screening against proteins without the prerequisite of a clean on-resonance frequency or defined ligand library. The new experiment can be used as the basis for studying intermolecular interactions where the standard STD experiment is difficult to optimize.