A natural consequence of groundwater withdrawals is the removal of water from subsurface storage, but the overall rates and magnitude of groundwater depletion and capture relative to groundwater withdrawals (extraction or pumpage) have not previously been well characterized. This study assesses the partitioning of long-term cumulative withdrawal volumes into fractions derived from storage depletion and capture, where capture includes both increases in recharge and decreases in discharge. Numerical simulation of a hypothetical groundwater basin is used to further illustrate some of Theis' (1940) principles, particularly when capture is constrained by insufficient available water. Most prior studies of depletion and capture have assumed that capture is unconstrained through boundary conditions that yield linear responses. Examination of real systems indicates that capture and depletion fractions are highly variable in time and space. For a large sample of long-developed groundwater systems, the depletion fraction averages about 0.15 and the capture fraction averages about 0.85 based on cumulative volumes. Higher depletion fractions tend to occur in more arid regions, but the variation is high and the correlation coefficient between average annual precipitation and depletion fraction for individual systems is only 0.40. Because 85% of long-term pumpage is derived from capture in these real systems, capture must be recognized as a critical factor in assessing water budgets, groundwater storage depletion, and sustainability of groundwater development. Most capture translates into streamflow depletion, so it can detrimentally impact ecosystems.