In a recently reported study [Mukherjee, et al. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 3528], we used 2D IR spectroscopy and 1-13C= 18O isotope labeling to measure the vibrational dynamics of 11 amide I modes in the CD3ζ£ transmembrane domain. We found that the homogeneous line widths and population relaxation times were all nearly identical, but that the amount of inhomogeneous broadening correlated with the position of the amide group inside the membrane. In this study, we use molecular dynamics simulations to investigate the structural and dynamical origins of these experimental observations. We use two models to convert the simulations to frequency trajectories from which the mean frequencies, standard deviations, frequency correlation functions, and 2D IR spectra are calculated. Model 1 correlates the hydrogen-bond length to the amide I frequency, whereas model 2 uses an ab initio-based electrostatic model. We find that the structural distributions of the peptidic groups and their environment are reflected in the vibrational dynamics of the amide I modes. Environmental forces from the water and lipid headgroups partially denature the helices, shifting the infrared frequencies and creating larger inhomogeneous distributions for residues near the ends. The least inhomogeneously broadened residues are those located in the middle of the membrane where environmental electrostatic forces are weakest and the helices are most ordered. Comparison of the simulations to experiment confirms that the amide I modes near the C-terminal are larger than at the N-terminal because of the asymmetric structure of the peptide bundle in the membrane. The comparison also reveals that residues at a kink in the α-helices have broader line widths than more helical parts of the peptide because the peptide backbone at the kink exhibits a larger amount of structural disorder. Taken together, the simulations and experiments reveal that infrared line shapes are sensitive probes of membrane protein structural and environmental heterogeneity.