The present paper reports Monte Carlo simulations results calculated at room temperature for a variety of aqueous systems, whose interaction potentials had been obtained by fitting quantum-mechanical ab-initio calculations, at CI (water–water) potential [1]) or SCF-LCAO-MO level [2–]. Unless otherwise stated, our systems were clusters consisting of one solute molecule and 200–250 water molecules, in vacuo; the pairwise additivity approximation was also used. 1) We simulated the two systems glycine–water (neutral molecule and zwitterion) [5]; their solvation structure was elucidated by means of orientational correlation functions (OCF's) and radial distribution factors (RDF's) [5]. The picture obtained by these means was found to be in broad, qualitative agreement with the information one can gain from (orientationally-optimized solute–water) isopotential contour maps. 2) We simulated the systems serine–water (neutral molecule and two conformers of the zwitterion) [6] and calculated RDF's and OCF's, hydrogen- and oxygen-atom probability density (PD) maps, and also investigated the spacial dependence of the average water–water and solute–water interaction energies. We also calculated PD maps for glycine zwitterion, and all the corresponding isopotential contour maps. Some water molecules are strongly hydrogen-bonded to hydrophilic groups (not to the alcoholic one) and, beyond them, there exists a rather rich and complex hydrogen-bonded network, resulting by a subtle balance between solvent–solvent and solvent–solute interactions, which is rather sensitive to conformational effects. 3) We also simulated a system consisting of an Na + cation, glycine zwitterion and water. Thermodynamic and structural results showed the Na + to be localized in the same plane as the −COO − group and close to it; no water molecules were found in the neighbourhood of the COO −⋯Na + pair, whereas the solvation of the amino group was not dramatically altered [7, 8]. 4) One could argue that a cluster in vacuo is not a very good approximation to a dilute solution; on the other hand, usage of periodic boundary conditions would probably require an even larger number of solvent molecules, and possibly an Ewald–Kornfeld summation for electrostatic potentials. As an alternative approach, we decided to consider our cluster as contained in spherical cavity [9] surrounded by a homogeneous and isotropic dielectric continuum, with the same dielectric constant as a pure solvent. The potential energy of the cluster due to the ‘dielectric reaction’ was evaluated on two serine zwitterion conformers showed a moderate increase (about 5%) in the solvation energy [10]. 5) In order to test the validity of the pairwise additivity approximation, quantum-mechanical calculations were carried out on the Li +(H 2O) 2 system, and two non-pairwise-additive three-body potential functions were extracted from the total interaction energies [11]. One of them appeared to be a short-range repulsive correction; the other one could be interpreted classically, as due to water bond polarization. Both terms were suitably generalized to a system Li + (H 2O) n, and Monte Carlo simulation was carried out on clusters with n ≤ 6, both with and without the many-body correction [11]. Comparison with available measured [12] and simulation [13] results showed a significantly improved agreement with experiment.