A specific method, combining some ingredients of the well-known DDA and PDI approaches, has been developped in our group since many years to calculate the absorption cross-sections of carbonaceous nanoparticles based on their atomistic details. This method, here named the Dynamic Atomic Dipole Interaction (DADI) model, requires the knowledge of the position and frequency-dependent polarizability of each atom constituting the nanoparticles. While the atomic positions can be quite easily obtained, for example as the results of molecular dynamics simulations, obtaining the frequency-dependent atomic polarizabilities is a trickier task. Here, a fitting procedure, named the reverse-DADI method, has been applied to calculate the frequency-dependent atomic polarizability values for carbon and hydrogen atoms involved in aromatic cycles or in aliphatic chains, on the basis of frequency-dependent molecular polarizabilities of various PAH and alkane molecules, calculated with the TD-DFT theory, in the UV–Visible range. Then, using these frequency-dependent atomic polarizabilities as input parameters in the DADI model has been shown to lead to an accurate representation of the absorption cross-sections of various PAH and alkane molecules with respect to the corresponding values obtained at the TD-DFT level, with however the great advantage of a much shorter time of calculations. Furthermore, these results are indications of a good transferability of the frequency-dependent atomic polarizability values obtained her e to any C or H atom of any PAH or alkane molecule. This opens the way for building large databases of optical properties for carbonaceous species of atmospheric or astrophysical interests.