In light of growing concern surrounding rising levels of atmospheric carbon dioxide (CO
2), there has been a great deal of renewed interest in alkanolamine-based chemisorption processes as part of the effort to mitigate anthropogenic emissions, particularly from large fixed-point emission sources such as fossil-fuel fired power-stations. The large-scale deployment of alkanolamine-based CO
2 capture technology will entail a significant economic penalty, and the reduction of this cost is a well-recognized imperative. A key element for CO
2 capture is the design of solvents and solvent blends with thermodynamic properties which would allow optimum capture operation processes. To describe and predict the thermodynamic properties of such solvents we have developed molecular models for use with the statistical associating fluid theory (SAFT). Specifically, the purpose is to identify mixtures that are good candidates for CO
2 absorption using transferable models and a group contribution framework within SAFT. In the proposed approach chemical equilibrium is treated with a physical approach, so that reaction products are treated implicitly. Two versions of the SAFT approach will be discussed; the SAFT-VR (SAFT for potentials of variable range) and the SAFT-γ (group contribution) equations of state. In particular, we present the theoretical basis of the applicability of SAFT-VR to carbon dioxide, water and alkanolamine mixtures, and discuss how SAFT-γ opens avenues for an improved thermodynamic prediction. The aqueous solvents examined include monoethanolamine (MEA), diethanolamine (DEA), methyldiethanolamine (MDEA), 2-amino-2-methyl-1-propanol (AMP), and their mixtures with CO
2. An excellent overall predictive capability is demonstrated for the fluid phase behaviour of pure components and binary aqueous mixtures. A good overall description is also obtained for the adsoprtion of CO2 in aqueous mixtures of alkanolamines, particulary for DEA and MDEA. The implicit treatment of the reaction equilibrium is validated by comparing predicted concentrations of reaction speciation products (carbamate and bicarbonate) with available experimental data.