In view of the controversy over climate change and variability, the last two decades have witnessed increased interest in the study of regional- and continental-scale trends in components of the terrestrial water cycle. In general, precipitation has increased significantly across the contiguous United States. An increase in precipitation will typically result in a greater relative increase in runoff. Several studies reported increasing trends of streamflow over the contiguous United States. Most of previous studies on trends in the water cycle relied on observational data and discussed trends at regional and continental scales. Moreover, most of these studies relied on station data that is not uniformly distributed in space and time and suffered from missing records. Evapotranspiration is typically computed as the difference between basin-averaged precipitation and streamflow at annual timescales. However, most major streams in the US are affected by regulation for water resources management purposes which affects the water budget computations. Physically-based computation of the water budget components through distributed model simulations provides a viable alternative to study the spatial distribution of natural trends in water cycle. Historical trends in the water budget components over the Southern Great Plains region of the United States were evaluated using land surface model simulations driven by observational forcings. Water and energy fluxes over the entire Arkansas/Red (AR) basin were simulated for the 1949-2000 period using the fully distributed land surface model. Annual and seasonal trends were studies with focus on the spatial distribution of the trends. Relationships among trends in different components of the water cycle were analyzed.