Quantifying changes in the terrestrial water cycle is central to understanding the availability of water resources, hydrological extremes, and the interactions between the climate and hydrological systems. Here the previous studies on the variability of terrestrial water budgets are comprehensively reviewed and synthesized. Overall, the observed variability in most of the hydrological variables is consistent with an intensification of the water cycle from regional to global scales. Under global warming, the frequency and intensity of precipitation have been found to increase over many parts of the world. Global annual evapotranspiration (ET) has seen an upward trend for 1982–1997 over land, followed by a decline trend afterward. However, substantial uncertainties still exist when we examine the retrospective hydroclimatic variability. The long-term trends for global precipitation and river discharges are still ambiguous. It is also inconclusive for the change in global land aridity. The validity of the “dry gets drier, wet gets wetter” pattern is prone to considerable uncertainties over land due to the lack of conclusive observational evidence. Our efforts suggest that it is of critical importance for the research community to continue improving the quantification of land surface hydrological fluxes and states toward a more reliable assessment of the terrestrial water cycle changes.