Due to new challenges, new extraction solvents based on innovative extractants are needed in hydrometallurgy for specific tasks. Thus, the aim of the present article is to discuss the potential and limits of Quantitative Structure-Properties Relationship (QSPR) and molecular modeling for identifying new extractants. QSPR methods may have useful applications in such a complex problem as the design of ligands for metal separation. Nevertheless, the degree of reliability of the predictions is still limited and, in the present state of the art, these techniques are likely more useful for optimization within a given family of extractants than to build in-silico new reagents. The molecular modeling techniques provide binding energies between target metals and given ligands, as well as optimized chemical structures of the formed complexes. Thus, in principle, the information, which can be deduced from the molecular modeling computations are richer than that provided by QSPR methods. Nevertheless, an effort should be made to establish more tangible links between the calculated binding energies and the physical parameters used by the hydrometallurgists, such as the complexation constants in aqueous phase (β MAn ) or better the extraction constants (K ex ).