Intrinsically unstructured proteins play key biochemical roles in a vast range of normal and pathological processes. To study these systems, it is necessary to invoke an ensemble of rapidly interconverting conformations. Residual dipolar couplings (RDCs) are particularly powerful probes of the behavior of unfolded proteins, reporting on time and ensemble-averaged conformations up to and beyond the millisecond time scale. In this study, we present a novel interpretation of RDCs in unfolded systems that simultaneously defines long-range structural order and local conformational sampling. This approach is used to describe the structure and dynamics of alpha-Synuclein (alphaS), a protein that is strongly implicated in the development of Parkinson's disease (PD), allowing unambiguous detection of strongly populated conformers containing long-range contacts between the N- and C-terminal domains. The structural model combines two features required for the description of alphaS in solution: local conformational fluctuation based on random sampling of residue-specific phi/psi distributions, and long-range contacts induced by the presence of nonbonding interactions between domains that are distant in primary sequence. Both aspects are found to be necessary for the reproduction of the nonaveraged RDCs from alphaS. Although RDCs have previously been shown to report on local conformational preferences in unstructured proteins, this study shows the additional sensitivity of these measurements to the presence of long-range order in highly flexible systems.