Proton two-dimensional nuclear Overhauser enhancement (2 D NOE) spectra in the pure absorption phase were obtained at 500 MHz for [d(GGAATTCC)], in aqueous solution at a series of mixing times. The experimental data were analyzed by comparison with theoretical spectra calculated using the complete 70 x 70 relaxation matrix including all proton dipole-dipole interactions and spin diffusion [Keepers, J. W. &James, T. L. (1984) J . M a p . Reson. 57,404-4261. The theoretical spectra at each mixing time were calculated using two structures: a standard B-form DNA structure and an energy-minimized structure based on the similarity of the six internal residues of the title octamer with those of the dodecamer [d(CGCGAATTCGCG)],, for which the crystal structure has been determined. Neither the standard B-form nor the energy-minimized structure will yield theoretical 2 D NOE spectra which accurately reproduce all peak intensities in the experimental spectra. However, many features of the experimental spectra can be represented by both the B-form and the energy-minimized structure. Sequencedependent structural characteristics are manifest in the 2 D NOE spectra, in particular at the purine-pyrimidine junction as noted previously in the crystal structure. On the whole, the energy-minimized structure appears to yield theoretical 2 D NOE spectra which mimic many, if not all, aspects of the experimental spectra. All 2 D NOE data were consistent with nanosecond correction times as implied by proton spin-lattice relaxation time measurements. But better fits of some of the 2 D NOE data using small variations in an effective isotropic correlation time suggest that there may be some local variations in mobility within the octamer duplex structure in solution.With recent advances in DNA synthesis enabling the preparation of DNA oligomers of known, specific sequence [l -41, X-ray crystallographers are examining DNA oligomers of different sequences and reporting details of sequence-dependent crystalline conformations [3, 5 -81. Questions regarding the structure of these DNA fragments in solution are also being addressed. These questions focus on the extent of preservation of the crystal structures in solution and on the manner in which the solution and crystal structures differ. On a more general level, discussion also focusses on whether or not it is accurate to speak of a molecular structure in solution and by what means may such a structure be determined. The continuing development of high-resolution NMR technology has both spurred such questions and been of great use in the quest for answers. Two-dimensional NMR (2D NMR), in particular, has developed to the point that proton spectra of biomolecules, spectra with perhaps hundreds of lines and once thought to be undecipherable, can now, in many cases, be totally assigned [I, 4,9 -111. Following assignment, both oneand two-dimensional NMR techniques enable spectroscopists to start probing in detail the solution conformations of these molecules.Correspondence to T. L. James, Departmen...