One potential means to decrease the level of atmospheric carbon dioxide is through the utilization of protein-CO(2) interactions. A recent bioinformatics analysis [Cundari TR et al. (2009) J Chem Inf Model 49:2111-2115] of these interactions revealed a marked disparity in CO(2) affinity between α-helices and β-sheets. In order to understand this difference, a series of molecular dynamics simulations was performed on polypeptide model systems. Numerous factors that may influence CO(2) affinity were systematically investigated, including the specific location of the amino acids within the secondary structural elements (SSEs), the partial charges on CO(2), chemical modifications made to the protein backbone, the inclusion of singly, doubly, and many functionalized residues, and the effect of solvent water. The differing abilities of the secondary structure types to participate in hydrogen bonding along the backbone were identified as being a crucial influence on CO(2) affinity; the lesser role of polypeptide-CO(2) electrostatic interactions was also noted. The effect of incorporating functionalized amino acid side chains, such as those possessed by Arg and His, on the affinity differs between the two structure types, and also strongly depends on the number included and the distance between them. The inclusion of explicit water molecules was found to attenuate all interactions, but did not change the overall trends in CO(2) affinity. Collectively, these results highlight the role of the backbone atoms in binding the CO(2) ligand, which will have important implications for efforts to ameliorate atmospheric carbon dioxide through the use of natural, designed, and modified proteins.