The amount and spatial distribution of deep drainage (downward movement of water across the bottom of the root zone) and groundwater recharge affect the quantity and quality of increasingly limited groundwater in arid and semiarid regions. We synthesize research from the fields of ecology and hydrology to address the issue of deep drainage in arid and semiarid regions. We start with a recently developed hydrological model that accurately simulates soil water potential and geochemical profiles measured in thick (>50 m), unconsolidated vadose zones. Model results indicate that, since the climate change that marked the onset of the Holocene period 10 000–15 000 years ago, there has been no deep drainage in vegetated interdrainage areas and that continuous, relatively low (<−1 MPa) soil water potentials have been maintained at depths of 2–3 m. A conceptual model consistent with these results proposes that the native, xeric‐shrub‐dominated, plant communities that gained dominance during the Holocene generated and maintained these conditions. We present three lines of ecological evidence that support the conceptual model. First, xeric shrubs have sufficiently deep rooting systems with low extraction limits to generate the modeled conditions. Second, the characteristic deep‐rooted soil–plant systems store sufficient water to effectively buffer deep soil from climatic fluctuations in these dry environments, allowing stable conditions to persist for long periods of time. And third, adaptations resulting in deep, low‐extraction‐limit rooting systems confer significant advantages to xeric shrubs in arid and semiarid environments. We then consider conditions in arid and semiarid regions in which the conceptual model may not apply, leading to the expectation that portions of many arid and semiarid watersheds supply some deep drainage. Further ecohydrologic research is required to elucidate critical climatic and edaphic thresholds, evaluate the role of important physiological processes (such as hydraulic redistribution), and evaluate the role of deep roots in terms of carbon costs, nutrient uptake, and whole‐plant development.