A high‐precision solution structure of the elastase inhibitor eglin c was determined by NMR and distance geometry calculations. A large set of 947 nuclear Overhauser (NOE) distance constraints was identified, 417 of which were quantified from two‐dimensional NOE spectra at short mixing times. In addition, a large number of homonuclear 1H‐1H and heteronuclear 1H‐15N vicinal coupling constants were used, and constraints on 42 χ1 and 38 ϕ angles were obtained. Structure calculations were carried out using the distance geometry program DG‐II. These calculations had a high convergence rate, in that 66 out of 75 calculations converged with maximum residual NOE violations ranging from 0.17 Å to 0.47 Å. The spread of the structures was characterized with average root mean square deviations () between the structures and a mean structure. To calculate the unbiased toward any single structure, a new procedure was used for structure alignment. A canonical structure was calculated from the mean distances, and all structures were aligned relative to that. Furthermore, an angular order parameter S was defined and used to characterize the spread of structures in torsion angle space. To obtain an accurate estimate of the precision of the structure, the number of calculations was increased until the and the angular order parameters stabilized. This was achieved after approximately 40 calculations. The structure consists of a well‐defined core whose backbone deviates from the canonical structure ca. 0.4 Å, a disordered N‐terminal heptapeptide whose backbone deviates by 0.8–12 Å, and a proteinase‐binding loop whose backbone deviates up to 3.0 Å. Analysis of the angular order parameters and inspection of the structures indicates that a hinge‐bending motion of the binding loop may occur in solution. Secondary structures were analyzed by comparison of dihedral angle patterns. The high precision of the structure allows one to identify subtle differences with four crystal structures of eglin c determined in complexes with proteinases. Copyright © 1992 The Protein Society