The study of the spatial sampling of biomolecules is essential to understanding the structure-dynamics-function relationship. We have established a protocol for the determination of multiple-state ensembles based on exact measurements of the nuclear Overhauser effect (eNOE). The protocol is practical since it does not require any additional data, while all other NMR data sets must be supplemented by NOE restraints. The question arises as to how much structural and dynamics information is shared between the eNOEs and other NMR probes. We compile one of the largest and most diverse NMR data sets of a protein to date consisting of eNOEs, RDCs and J couplings for GB3. We show that the eNOEs improve the back-prediction of RDCs and J couplings, either upon use of more than one state, or in comparison to conventional NOEs. Our findings indicate that the eNOE data is self-consistent, consistent with other data, and that the structural representation with multiple states is warranted. (C) 2015 Elsevier Inc. All rights reserved.