Abstract
A resolution to the controversy over the magnitude of the effective short-range Coulomb interaction in molecular conductors is proposed by showing that it depends very strongly on band filling because of intermolecular correlations.