Carbon dioxide (CO2) capturing has been an important issue for decades. Alkanoamines, such as diethanolamine (DEA) have been widely used for CO2 separation by absorption process. During this process, CO2 loading measurement is an imperative action for a proper process control. Currently used methods are titration based which requires a long processing time. In this work Raman spectroscopy is used to model and predict the CO2 loading in wide range (0-0.97 CO2 mole/amine mole). The models are developed by using Raman peak ratios to minimize the error due to peaks fluctuations. The Raman peak ratio of 1022 cm-1/1461cm-1 has been found as a good fit with the coefficient of determination (R2) of 0.92 and mean squared error (MSE) of 0.00656 CO2 mole2/ amine mole2 in prediction of CO2 loading. © 2014 Trans Tech Publications, Switzerland.