Tropical forest degradation (caused by activities like selective logging, fires, and fragmentation) is as widespread as deforestation. However, little is known about how structural and functional changes in degraded forests affects the energy, water, and carbon cycles at regional scale. USDA Forest Service researchers and partners used the Ecosystem Demography Model (ED-2.2) to investigate how degradation-driven forest structure affects sensible heat, evapotranspiration and gross primary productivity. To collect the detailed forest structure data for the model, small-footprint airborne laser scanning (lidar) data calibrated with a large number of forest inventory plots across precipitation and degradation gradients in the Amazon was used. Results show that tropical forest degradation effects were the strongest in seasonal forests and during near-climatological (normal) years. Increased water stress in degraded forests resulted in up to 35% reduction in evapotranspiration and gross primary productivity, and up to 43% increase in sensible heat flux. Degradation effects diminished during extreme droughts, when water stress dominated the response even in intact forests. These results indicate a broad influence of land cover and land use change in energy, water, and carbon cycles that is not limited to deforested areas, and highlight the relevance of structural diversity in predicting biophysical and biogeochemical cycles.