The configurational analysis of organic compounds is an important application for high resolution NMR spectroscopy. In the present study, a tetra-substituted pyrrolidine with four chiral carbon atoms is analyzed using classical methods based on 3J and NOE data in solution and compared and verified with recently introduced alternative approaches via residual dipolar couplings (RDCs) in two weak anisotropic alignment media. The molecule shows sufficient rigidity in the five-membered ring for the configurational characterization with the various techniques. However, the flexibility caused by the many freely rotating bonds potentially poses problems for the interpretation of data. It is shown that RDCs measured in poly-γ-benzyl-l-glutamate and a stretched polydimethylsiloxane gel provide useful information for the distinction of diastereomers, but the success varies with the data interpretation strategy used. Although a general improvement of corresponding correlation factors is observed when limiting data to a subset of dipolar couplings directly connected to the central ring, the distinction power is reduced because of the smaller number of RDCs available for potential model falsification. Singular value decomposition for fitting experimental RDCs is able to distinguish in most cases the correct from incorrect configurations, but the differences in correlation factors can be relatively small. Surprisingly, predicting RDCs using the rod model as implemented in PALES gives best results in distinguishing the eight possible diastereomers. It is also found that the use of proton–phosphorus and carbon–phosphorus RDCs helps with the configurational analysis of the model compound. Copyright © 2012 John Wiley & Sons, Ltd.