Although functionalization is known as a promising way to notably improve the performances of ionic liquids (ILs) in separation processes, studies thereon are mainly based on experimental trial and error while rational design of functional ILs is scarcely reported. In this work, computer-aided IL design (CAILD) and molecular dynamics (MD) simulation are combined toward rationally functionalizing ILs for the enhanced extractive desulfurization (EDS) of fuel oils. First, the UNIFAC-IL model is extended based on experimental data to specifically cover interaction parameters associated with four functional groups (i.e., hydroxyl, methoxy, vinyl, and cyanomethyl). A mixed-integer nonlinear programming problem is then formulated for the computer-aided design of functional ILs for the EDS task. Finally, MD simulations are performed to study the intermolecular interactions between the top IL candidates and model fuel oil components; compared with those for literature-reported ILs, the enhanced EDS performances of the designed functional ILs are well rationalized.