The influence of the extractive solvent/hydrocarbons feed ratio, reflux ratio, and column configuration on the separation of aromatics from a hexane, heptane, cyclohexane, benzene, and toluene mixture using N,N-dimethylformamide was studied both experimentaly and by simulation, using a commercial process simulator, phase equilibria predictions by the models NRTL, Wilson, and UNIQUAC, and model parameters available in the literature. The results show that the efficiency for the separation of aromatic/non-aromatic hydrocarbon mixtures increases when the solvent/hydrocarbon feed ratio and the number of column plates are increased but is unaffected by the reflux ratio within the experimental range. Simulations predict qualitatively the influence of the operation variables on the column performance but are not accurate enough for design purposes.