A series of acene-modified zinc-porphyrin dyes (benzene to pentacene, denoted as LAC-1 to were chosen to examine their performance as photosensitizers in dye-sensitized solar cells (DSSCs). Their structural, electronic and optical properties were investigated at the DFT/TDDFT levels using various theoretical models (i.e., the gas phase model and the implicit/explicit solvent model). The Dye@TiO 2 complex was used to investigate the dye/semiconductor interfaces using both the cluster and periodic models. After a careful examination of the dependence of the results on different theoretical approaches, some basic principles could be derived based on the theoretical investigation of structure-function relationships in isolated dyes and dye-TiO 2 assemblies. Based on these ideas, some general suggestions can be proposed for the future design of dyes for use in DSSCs. For instance, the DFT functionals used in estimating the critical parameters for DSSCs should be carefully validated.Sometimes the performances of the DFT functionals can be improved by a specific energy-shift correction to compensate for systematic errors. Benchmark calculations indicated that the best approach for depicting the reduction potentials is either the M06-2X functional combined with the formula ∆E red = (E 0 -E -) GS or the B3LYP functional combined with Koopman's Theorem. The best functional for estimating the excitation energies was found to be LC-ωPBE. The impact of significant thermal fluctuations on the optoelectronic properties of dyes may also be an important consideration in the prediction of more efficient dyes for use DSSCs. In contrast to the selection of DFT functionals, both the cluster and periodic models resulted in consistent views of the dye-TiO 2 interactions, indicating that the use of either model should achieve reasonable results at least in the qualitative manner.