Climate and other environmental conditions have varied in the past, and will almost certainly vary significantly in the near future. The response of groundwater recharge to changes in environmental conditions is thus a matter of active concem for water-resources management. The major mechanisms for this response of recharge are three-fold. First, changes in vegetation communities can shift the water balance at the base of the root zone, increasing or decreasing the amount of recharge. Second, variations in the amount of runoff can affect channel recharge. Finally, shifts in the seasonality of precipitation can strongly affect the fraction that is evapotranspired back into the atmosphere and thus affect the amount that becomes recharge. Increases in recharge (defined as the water flux across the water table) may in some cases significantly increase fluxes through regional aquifers, but in other cases, depending on the hydrogeology, may only result in increased streamflow or evapotranspiration within the recharge area. Basins with relatively low maximum elevations, deep water tables, thin soils, and highly permeable recharge areas experience the largest recharge response to increases in precipitation. The relatively well-known paleoenvironmental history of the American Southwest can be compared with various lines of evidence for changes in recharge. These lines of evidence include timing of speleothem formation, chloride profiles in thick vadose zones, changes in water table shown by subsurface calcite precipitation, and expanded groundwater discharge areas. This evidence indicates that the wettest periods of the past 25 ka, which were generally between 20 and 13 ka, were also periods of enhanced vadose zone fluxes and aquifer discharge. Climate-driven changes in recharge appear to have been substantially mediated through changes in vegetation. This evidence for strong recharge response to past environmental changes indicates that expected future climate and environmental change will also cause changes in recharge. The ability to adequately predict future changes in recharge will depend on developing process-based numerical models that can simulate coupled climate/vegetation/ vadose zone processes and incorporate the outputs into groundwater/surface water models that can resolve processes at scales ranging from the hillslope to the basin.