The increasing importance of natural gas as an energy source poses separation challenges, due to the high pressures and high carbon dioxide concentrations of many natural gas streams. A methodology for computer-aided molecular and process design (CAMPD) applicable to such extreme conditions is presented, based on the integration of process and cost models with an advanced molecular-based equation of state, the statistical associating fluid theory for potentials of variable range (SAFT-VR). The approach is applied to carbon dioxide capture from methane using physical absorption. The search for an optimal solvent is focused on n-alkane blends. A simple flowsheet is optimised using two objectives: maximum purity and maximum net present value. The best equipment sizes, operating conditions, and average chain length of the solvent (the n-alkane) are identified, indicating n-alkane solvents offer a promising alternative. The proposed methodology can readily be extended to wider classes of solvents and to other challenging processes. (c) 2010 Elsevier Ltd. All rights reserved.