The magnitude of hydraulic redistribution of soil water by roots and its impact on soil water balance were estimated by monitoring time courses of soil water status at multiple depths and root sap flow under drought conditions in a dry ponderosa pine (Pinus ponderosa Dougl. ex Laws) ecosystem and in a moist Douglas-fir (Pseudotsuga menziesii (Mirb.) Franco) ecosystem. The fate of deuterated water applied to small plots to create a strong horizontal soil water potential gradient was also monitored to assess the potential for horizontal redistribution of water and utilization of redistributed water by co-occurring shallow-rooted plants. In a 20-year-old Douglas-fir stand, approximately 28% of the water removed daily from the upper 2 m of soil was replaced by nocturnal hydraulic redistribution during late August. In an old-growth ponderosa pine stand, approximately 35% of the total daily water utilization from the upper 2 m of soil appeared to be replaced by hydraulic redistribution during July and August. By late September, hydraulic redistribution in the ponderosa pine stand was no longer apparent, even though total water use from the upper 2 m of soil was nearly identical to that observed earlier. Based on these results, hydraulic redistribution would allow 21 and 16 additional days of stored water to remain in the upper soil horizons in the ponderosa pine and Douglas-fir stands, respectively, after a 60-day drought. At both sites, localized applications of deuterated water induced strong reversal of root sap flow and caused soil water content to cease declining or even temporarily increase at locations too distant from the site of water application to have been influenced by movement of water through the soil without facilitation by roots. Xylem water deuterium values of ponderosa pine seedlings suggested utilization of redistributed water. Therefore, hydraulic redistribution may enhance seedling survival and maintain overstory transpiration during summer drought. These first approximations of the extent of hydraulic redistribution in these ecosystems suggest that it is likely to be an important process in both wet and dry forests of the Pacific Northwest.