The quantum chemical Conductor-like Screening Model for Real Solvents (COSMO-RS) method was evaluated as a theoretical framework to computationally investigate the application of room temperature ionic liquids (ILs) in absorptive technologies for capturing CO2 from power plant emissions to efficiently reduce both experimental efforts and time consumption. First, different molecular models to simulate ILs and computational methods in geometry calculations were investigated to optimize the COSMO-RS capability to predict Henry's Law coefficients using a demanding solubility sample test with 35 gaseous solute-IL systems and 20 CO2-IL systems. The simulation results were in good agreement with experimental data, indicating that using an ion-pair molecular model optimized in a gas-phase environment allows a finer COSMO-RS description of the IL structure influence on the CO2 and other solutes solubilities. Moreover, the COSMO-RS methodology was used for the first time to achieve a deeper insight into the behavior of the solubility of CO2 in Its from a molecular point of view. For this purpose, further analyses of the energetic intermolecular interactions between CO2 and ILs were performed by COSMO-RS, revealing that the van der Waals forces associated with the solute in the liquid phase determine the absorption capacity of CO2 in Its, which is measured in terms of Henry's Law coefficients. These findings were finally driven by a rational screening over 170 as with COSMO-RS to design new ILs that enhance CO2 capture by physical absorption.