Accurate estimation of ecosystem-scale land surface energy and water balance has great importance in weather and climate studies. This paper summarizes seasonal and interannual fluctuations of energy and water balance components in two distinctive semiarid ecosystems, sagebrush (SB) and lodgepole pine (LP) in the Snake River basin of Idaho. This study includes 6 years (2011–2016) of eddy covariance (EC) along with modeled estimates. An analysis of the energy balance indicated a higher energy balance ratio (0.88) for SB than for LP (0.86). The inclusion of canopy storage (CS) increased the association between turbulent fluxes and available energy in LP. Green vegetation fraction (GVF) significantly controlled evapotranspiration (ET) and surface energy partitioning when available energy and soil moisture were not limited. Seasonal water balance in the Budyko framework showed severe water-limited conditions in SB (6–9 months) compared to LP (6–7 months). Based on the validated Noah land surface model estimates, direct soil evaporation (ESoil) is the main component of ET (62 to 79%) in SB due to a large proportion of bare soil (60%), whereas at the lodgepole pine site, it was transpiration (ETran, 42–52%). A complementary ratio (CR) analysis on ET and potential ET (PET) showed a strong asymmetric CR in SB, indicating significant advection. Both SG and LP showed strong coupling between soil moisture (SM) and air temperature (Ta). However, a weak coupling was observed in SB when the soil was dry and Ta was high. This weak coupling was due to the presence of net advection. The results presented here have a wider application: to help us understand and predict the survival, productivity, and hydroclimatology of water-limited ecosystems.