The observation that in efficient organic solar cells almost all electron–hole pairs generated at the donor–acceptor interface escape from their mutual coulomb potential remains to be a conceptual challenge. It has been argued that it is the excess energy dissipated in the course of electron or hole transfer at the interface that assists this escape process. The current work demonstrates that this concept is unnecessary to explain the field dependence of electron–hole dissociation. It is based upon the formalism developed by Arkhipov and co-workers as well as Baranovskii and co-workers. The key idea is that the binding energy of the dissociating “cold” charge-transfer state is reduced by delocalization of the hole along the polymer chain, quantified in terms of an “effective mass”, as well as the fractional strength of dipoles existent at the interface in the dark. By covering a broad parameter space, we determine the conditions for efficient electron–hole dissociation. Spectroscopy of the charge-transfer state on bilayer solar cells as well as measurements of the field dependence of the dissociation yield over a broad temperature range support the theoretical predictions.