Abstract An attempt is made to calculate the second-derivative contribution to the electroabsorption signal of the anthracene crystal. A model proposed previously for the description of charge transfer (CT) states is modified to include some effects of the coupling of excitons with intramolecular vibrations. It is argued that only the CT states from the edge of the Brillouin zone ( k= ±π) can contribute to electroabsorption, and even those are forbidden unless accompanied by phonons. The results of such an approach are in good agreement with experiment, which can be further improved in the future by ameliorating some approximations of the treatment. They seem to indicate that the “optical” band gap in anthracene is 4.25 eV, rather than 4.42 eV as supposed previously in the literature. This considerably reduces the discrepancy between the “optical” and “electrical” band gaps of this crystal.