A predictive model was developed for corrosion of carbon steel based on modeling of concentrations of corrosive species and electrochemical reactions in CO2-loaded aqueous methyldiethanolamine (MDEA) systems. The concentrations of aqueous carbonic and amine species (CO2 HCO3- CO32- MDEA and MDEAH+) as well as pH values in the MDEA solution were calculated to provide a tool for estimating water chemistry using equilibrium constants for each chemical reaction at various CO2 partial pressures. The water chemistry model showed a good match with experimental data for pH and CO2 loading with improved correlations upon use of activity coefficients. The electrochemical corrosion model was developed by establishing polarization curves based on the determined species concentrations. The required electrochemical parameters (e.g. exchange current densities Tafel slopes and reaction orders) for different reactions were determined from experiments conducted in glass cells. Iron (Fe) oxidative dissolution bicarbonate (HCO3-) reduction and protonated alkanolamine (MDEAH+) reduction reactions were successfully simulated to build a corrosion model for MDEA-CO2-H2O environments. The model was applicable for uniform corrosion when no protective films are present. Furthermore this provides a solid foundation for corrosion model development for other amine-based CO2 capture processes.