Modeling and observational studies suggest that variations in the state of the land surface, including soil moisture and temperature, can have important implications for local weather and regional climate. We have examined land-atmosphere coupling over the continental United States in an ensemble of regional climate simulations developed in conjunction with the NA-CORDEX and DOE FACETS projects. This ensemble of regional climate simulations includes two different regional climate models (RCMs; WRF and RegCM4), driven with lateral-boundary conditions from three different GCMs. Downscaling experiments with each combination of RCM and GCM were performed at three different resolutions (12 km, 25 km, and 50 km) and run for historical and future climate time periods. Past studies of land-atmosphere coupling have used both models and observations, but to our knowledge this is the first comparative evaluation using a systematic ensemble of regional climate models. Multiple land-atmosphere coupling diagnostic metrics are used to assess regional and seasonal variations in land-atmosphere coupling within this RCM ensemble. Initial results show that there are systematic differences in land-atmosphere coupling strength between the RCMs examined in the study. We also find there are less pronounced differences corresponding with variations in the horizontal grid spacing and choice of model used to drive the RCMs. These differences in land-atmosphere coupling strength between the simulations provide insight into sources of model bias and drivers of future changes in precipitation and temperature.