Several approaches for utilizing dipolar recoupling solid-state NMR (ssNMR) techniques todetermine local structure at high resolution in peptides and proteins have been developed. However, manyof these techniques measure only one torsion angle or are accurate for only certain classes of secondarystructure. Additionally, the efficiency with which these dipolar recoupling experiments suppress thedeleterious effects of chemical shift anisotropy (CSA) at high magnetic field strengths varies. Dipolarrecoupling with a windowless sequence (DRAWS) has proven to be an effective pulse sequence for excitingdouble-quantum (DQ) coherences between adjacent carbonyl carbons along the peptide backbone. Byallowing this DQ coherence to evolve, it is possible to measure the relative orientations of the CSA tensorsand subsequently use this information to determine the Ramachandran torsion angles
and
. Here, weexplore the accuracies of the assumptions made in interpreting DQ-DRAWS data and demonstrate theirfidelity in measuring torsion angles corresponding to a variety of secondary structures irrespective ofhydrogen-bonding patterns. It is shown how a simple choice of isotopic labels and experimental conditionsallows accurate measurement of backbone secondary structures without any prior knowledge. This approachis considerably more sensitive for determining structure in helices and has comparable accuracy for
-sheetand extended conformations relative to other methods. We also illustrate the ability of DQ-DRAWS todistinguish between structures in heterogeneous samples.