A short linear peptide in solution may populate several stable states (denoted here microstates) in thermodynamic equilibrium. Elucidating its dynamic 3D structure by multidimensional nuclear magnetic resonance (NMR) is complex, since the experimentally measured nuclear Overhauser effect intensities (NOEs) represent averages over the individual contributions. In previous papers (paper 1, Meirovitch, H.; et al. J. Phys. Chem. 1995, 99, 4847; and paper 2, Meirovitch, E.; Meirovitch, H. Biopolymers 1996, 38, 69) we have developed a new theoretical methodology based on statistical mechanical considerations for analyzing NMR data from flexible molecules and applied it to Leu-enkephalin (H-Tyr-Gly-Gly Phe-Leu-OH) using the potential energy function ECEPP. The corresponding experimental NOEs have been obtained from a cryoprotective solution of this molecule. Our approach is based on a conformational search which identifies a set of significantly different low-energy structures within a certain energy range above the global energy minimum. These structures are taken as "seeds" for Monte Carlo (MC) simulations which span their vicinity; the corresponding samples are called MC microstates. Their free energies, hence populations, are obtained by the local states (LS) method, and the individual contribution of each microstate to the NOEs is calculated. The overall NOE intensity is given by the average of these individual contributions, weighted by the populations of the MC microstates. Here we apply this methodology to the same molecule described by the ECEPP energy and a solvation free energy term for water developed by Wesson and Eisenberg. This term is a summation over products of the solvent-accessible surface area of each atom and its solvation parameter. Since water is the most important solvent in biological systems, investigating the properties of this model is an imperative step in the development of our methodology. Thus, the energy barriers of the solvation model are expected Co be lower than those of ECEPP alone; hence it is crucial to verify that the MC microstates of the former model are thermodynamically stable and structurally distinctive (i.e., they do not overlap). Criteria for these purposes, proposed in paper 2, are further developed here and applied to the MC microstates. We verify another crucial point, that converging results for the free energy are obtained with the LS method. In agreement with Leu-enkephalin in water being a random coil, we find that MC microstates of the solvation model have larger entropy and structural diversity than those based on ECEPP (paper 2), and they do not feature intramolecular hydrogen bonds. These physical and computational properties of the solvation model suggest that it would be applicable to other peptide systems as well. We also compare the NOEs predicted by the two models and outline plans for future work.