Using FIFE observations on four golden days and a coupled soil-canopy-atmosphere (mesoscale) dynamic model, we simulate the biophysical influences of a prairie on the surface energy budget. Atmospheric dynamics within the model are fully three-dimensional. Canopy and soil submodels were modified to include the aggregate effects of standing brown vegetation on stomatal conductance and surface energy partitioning, and litter cover on soil thermal and hydraulic properties. Model-simulated energy fluxes compare reasonably well to in situ measurements. It was found that the spatial variation of simulated latent-, sensible-, and ground-heat fluxes is very sensitive to soil-moisture and biomass distributions over the FIFE domain. High latent-heat fluxes are related to high occurrences of green leaf-area indices (LAIgreen), low brown leaf-area indices (LAIbrown), and high soil-moisture contents, as well as to sufficient available energy. High sensible-heat fluxes often are coincident with a low LAIgreen, high LAIbrown, and relatively dry soils. High ground-heat fluxes occur with lower values of total LAI and high soil-moisture contents. As a result of soil-canopy-atmosphere interactions, simulated air temperatures vary spatially up to 2degree C. In addition, it was found that-for a prairie environment like the Konza Prairie with green leaf-area indices less than 2-simulated latent-heat fluxes from the canopy are about two times greater than those from the soil beneath the canopy