As the issue of greenhouse gases, particularly carbon dioxide, becomes more prevalent in studying land-climate interactions, models that accurately represent vegetation are critically important since plants store and emit large amounts of CO2. Recent studies note that climate models fail to reproduce observational data of phenology. Evaluating models against observations helps identify needs and approaches for model improvement. The focus of this study is to evaluate the performance of the new crop option in the CLM4CN/CAM4 by comparing two sensitivities of the coupled simulation to observed data gathered from the Ameriflux network website. The Late-planting simulation prescribes planting crops as late as the conditions in the simulation allow, while the Crop simulation plants crops whenever the conditions allow during the planting period. The selected locations, Bondville, IL( US-Bo1), Ponca City OK (US-Pon), and Fermi, IL (US-IB1) revealed better accuracy using the Late-planting simulation. US-Bo1 revealed leaf area index (LAI) heavily influenced net ecosystem exchange (NEE) and sensible (H) and latent (L) heat fluxes. Increasing LAI results in an increase in L due to more transpiration from leaves. As a result less of the net radiation goes toward H. The timing of maximum LAI also influences carbon uptake and, therefore, the NEE. US-Pon exhibited the model's inability to accurately represent winter wheat while US-IB1 disclosed small influences from crops due to very small crop coverage in the grid cell. Further research at additional locations will help determine the accuracy of model representation of crop vegetation on land-climate interactions.