The Amazon River is the largest river by basin area and discharges by far the largest amount of fresh water to the ocean worldwide. Despite previous efforts based on observations and modeling, there remains large uncertainty in estimating the evapotranspiration and runoff in the Amazon Basin. This has important implications to climate modeling because the Amazon plays a key role in regulating the global water cycle. A team of researchers including scientists from Pacific Northwest National Laboratory used a suite of state-of-the-art land surface models (LSMs) to evaluate the water budget of the Amazon basin. Several sets of meteorological forcing datasets were used as inputs to 14 LSMs for comparison. At the basin scale, they compared water budget variables including terrestrial water storage, evapotranspiration, surface runoff, and base flow with both remote sensing and in-situ data. They compared river discharges simulated by a river routing scheme forced with runoff and base-flow simulated by the LSMs against observations at 165 gauges. Then, simulated evapotranspiration were compared against the flux tower and remotely sensed evapotranspiration estimates, and they used the Gravity Recovery and Climate Experiment (GRACE) terrestrial water storage estimates to evaluate the simulated terrestrial water storage in Madeira and Negro Rivers, two sub-catchments of the main tributaries. The simulated basin averaged evapotranspiration ranged from 2.39 to 3.26 mm/day. They found a low spatial correlation between evapotranspiration and precipitation indicating that evapotranspiration does not depend on water availability over most of the basin. Results also show that other simulated water budget components vary significantly as a function of both the LSM and precipitation dataset, but the simulated terrestrial water storage generally agrees with the GRACE estimates at the basin scale. The best water budget simulations resulted from the use of a precipitation dataset compiled from a denser rainfall gauge network and the rescaling at a finer temporal scale. While results generally showed that representing physical processes in a higher level of complexity improves simulation of hydrological processes, more analysis considering energetic variables in addition to hydrologic variables is needed to better understand differences in model skill.