Optimizing the optical properties of large chromophore aggregates and molecular solids for applications in photovoltaics and nonlinear optics is an outstanding challenge. It requires efficient and reliable computational models that must be validated against accurate theoretical methods. We show that linear absorption spectra calculated using the molecular exciton model agree well with spectra calculated using time-dependent density functional theory and configuration interaction singles for aggregates of strongly polar chromophores. Similar agreement is obtained for a hybrid functional (B3LYP), a long-range corrected hybrid functional (ωB97X), and configuration interaction singles. Accounting for the electrostatic environment of individual chromophores in the parametrization of the exciton model with the inclusion of atomic point charges significantly improves the agreement of the resulting spectra with those calculated using all-electron methods; different charge definitions (Mulliken and ChelpG) yield similar results. We find that there is a size-dependent error in the exciton model compared with all-electron methods, but for aggregates with more than six chromophores, the errors change slowly with the number of chromophores in the aggregate. Our results validate the use of the molecular exciton model for predicting the absorption spectra of bulk molecular solids; its formalism also allows straightforward extension to calculations of nonlinear optical response.