A new approach for the prediction of retention in reversed-phase liquid chromatography using ternary mobile phase compositions is presented. The solvation parameter model is used to generate system constants characteristic of the capability of the chromatographic system for defined solvent–solvent and solute–solvent interactions. A mixture-design approach is then used to construct system surfaces for each system constant over all mobile phase compositions. These surfaces are smooth and devoid of irregular features. Models derived from these surfaces allowed us to predict the retention of a varied group of 36 solutes on a porous polymer sorbent, PLRP-S 300, at a methanol–acetonitrile–water composition not included in the data used to construct the system surfaces. The accuracy of the predicted retention factors was similar to that obtained by fitting the solvation parameter model to the same experimental data.