Stable water isotopes have been employed as a means of challenging, validating, and improving numerical models of the Amazon Basin since the 1980s. This paper serves as an exemplar of how characterization of human and natural impacts on surface-atmosphere water exchanges could beneficially exploit stable water isotope data and simulations. Interpretations of Amazonian isotopic data and model simulations are found to be seriously hampered by (1) poor simulation of the gross water budget (e.g., lack of surface water conservation in models); (2) considerable model differences in the fate of precipitation (i.e., between reevaporation and runoff); (3) wide ranging characterization of natural causes of water isotopic fluctuations (especially El Ni¿o and La Ni¿a events); (4) isotopic land-atmosphere flux sensitivity to the prescription of boundary layer atmospheric water vapor isotopic depletion; and (5) significantly different characterization by current land-surface schemes of the partition of evaporation between isotopically fractionating (from lakes and rivers) and nonfractionating (transpiration) processes. Despite these obstacles, we find features in the recent isotopic record that might be derived from circulation and land-use changes. ENSO events may cause decreased depletion in the dry season, because of reduced convective precipitation, while increases in upper basin isotope depletions in the wet season may result from relatively less nonfractionating recycling because there are fewer trees. The promise for isotopic fingerprinting of near-surface continental water cycle changes depends upon fixing shortcomings in current atmospheric and land-surface models.