Ionic liquids (ILs) is an effective way to convert and remove hydrogen sulfide, H2S from natural gas streams. Amine-based ILs have recently been discovered due to their efficiency of removing H2S due to their excellent characteristics. However, the selection of potential amine-based ILs from a wide range combination of cations and amine-based anions are rather challenging if performed experimentally. Conductor-like Screening Model for Real Solvents (COSMO-RS) was used for the prediction of liquids thermodynamics such as activity coefficient at infinite dilution, selectivity, capacity and performance index (PI) values to evaluate the performance of each amine-based ILs and determine the most efficient ILs for H2S absorption without requiring any experimental data. A total of 42 previously unexplored ionic liquids, amine-based ILs from a combination of 7 amine cations and 6 anions were screened by COSMO-RS model. The findings from COSMO-RS prediction results revealed that ILs without aromatic rings displayed higher efficiencies compared to ILs with aromatic rings. Moreover, N,N,N�,N�-tetramethyl-1,3-propanediamine chloride, TMPDACl cation-anion combinations displayed the highest PI value of 52.55 among the screened ILs thus, predicted as the most efficient ILs for H2S absorption. The experimental results obtained validated the COSMO-RS predictions with an approximation of ± 4.51% discrepancies between predicted and experimental data. Hence, it is concluded that COSMO-RS can be used for a reliable pre-experimental prediction and amine-based ILs approach has a huge potential for effective H2S absorption in industrial applications. © 2024 Institution of Chemical Engineers