Chemical absorption of CO2 into aqueous amine-based solvents is known as the most mature technology featuring high separation efficiencies and applicability of retrofitting. The high regeneration energy requirement of the process is a major drawback, and improvements in solvent design are required. This study investigated the CO2 absorption and desorption performance of the nonaqueous monoethanolamine (MEA)/methyl diethanolamine (MDEA) blend experimentally using a stirred cell reactor. CO2 loading, cyclic capacity loss, and initial absorption rate were measured for different solvent formulations and compared to single amines (MEA or MDEA). A mixture-process design and response surface methodology were employed to model and optimize the solvent formulation at different temperatures and pressures. The single-response optimization yielded 0.83molCO2/molamine (at 0% MEA and 303 K/0.5 barg) and 3.17e-5 kmol/m(2)center dot s (at 40% MEA and 310.67 K/0.5 barg) as optimal absorption capacity and rate of absorption, respectively. The multiresponse optimization was conducted using the composite desirability function yielding 0.653molCO2/molamine and 2.987e-5 kmol/m(2)center dot s (D = 0.903). The multiresponse optimization was also extended to include the impact of different initial settings and importance ratios between the absorption capacity and rate of absorption, in which the latter needed at least 1.6 higher importance level to influence the multiresponse optimization.