this protocol is intended to provide chemists who discover or make new organic compounds with a valuable tool for validating the structural assignments of those new chemical entities. experimental 1 H and/or 13 c nMr spectral data and its proper interpretation for the compound of interest is required as a starting point. the approach involves the following steps: (i) using molecular mechanics calculations (with, e.g., MacroModel) to generate a library of conformers; (ii) using density functional theory (DFt) calculations (with, e.g., Gaussian 09) to determine optimal geometry, free energies and chemical shifts for each conformer; (iii) determining Boltzmann-weighted proton and carbon chemical shifts; and (iv) comparing the computed chemical shifts for two or more candidate structures with experimental data to determine the best fit. For a typical structure assignment of a small organic molecule (e.g., fewer than ~10 non-H atoms or up to ~180 a.m.u. and ~20 conformers), this protocol can be completed in ~2 h of active effort over a 2-d period; for more complex molecules (e.g., fewer than ~30 non-H atoms or up to ~500 a.m.u. and ~50 conformers), the protocol requires ~3-6 h of active effort over a 2-week period. to demonstrate the method, we have chosen the analysis of the cis-versus the trans-diastereoisomers of 3-methylcyclohexanol (1-cis versus 1-trans). the protocol is written in a manner that makes the computation of chemical shifts tractable for chemists who may otherwise have only rudimentary computational experience. this method certainly has value, the example described next shows that when one moves to the consideration of molecules bearing more than one stereocenter, this approach is no longer adequate.Consider the case of the trans-versus the cis-diastereomers of 3-methylcyclohexanol (1-trans and 1-cis, respectively). The experimental 1 H NMR spectra for 1-trans and 1-cis are shown in Figure 1a,b (see Supplementary Data 1 for a full listing of actual chemical shift values). There are substantial differences in these two spectra, especially within the upfield 0.7-2.1 p.p.m. range. Clearly, it would be valuable if computational approaches could reproduce these sorts of differences sufficiently well to allow confident assignment of structure.Common software packages that use empirical (often increment-based) compilations of chemical shift information (e.g., tabulated shift increments or databases of known spectral data) allow users to predict the chemical shifts of a given input structure. These include 'ChemNMR' within ChemBioDraw (also known as ChemDraw) and 'C+H NMR Predictor and DB' within the ACD/Labs software suite. These methods sometimes can be sufficient for the task of resolving constitutional structural assignments. However, when issues associated with relative configuration are considered, increment-based methods are decidedly ill-equipped. Analysis of structures 1-trans and 1-cis by each of these programs quickly reveals these limitations, even for these simple structures. Namely, because Che...