The use of plants to remediate polluted groundwater is becoming an attractive alternative to more expensive traditional techniques. In order to adequately assess the effectiveness of the phytoremediation treatment, a clear understanding of water-use habits by the selected plant species is essential. We examined the relative uptake of surface water (i.e., precipitation) vs. groundwater by mature Populus deltoides by applying irrigation water at a rate equivalent to a 5-cm rain event. We used stable isotopes of hydrogen (D) and oxygen (18O) to identify groundwater and surface water (irrigation water) in the xylem sap water. Pretreatment isotopic ratios of both deuterium and 18O, ranked from heaviest to lightest, were irrigation water > groundwater > xylem sap. The discrepancy in preirrigation isotopic signatures between groundwater and xylem sap suggests that in the absence of a surface source of water (i.e., between rain events) there is an unknown amount of water being extracted from sources other than groundwater (i.e., soil surface water). We examined changes in volumetric soil water content (%), total hourly sapflux rates, and trichloroethene (TCE) concentrations. Following the irrigation treatment, volumetric soil water increased by 86% and sapflux increased by as much as 61%. Isotopic signatures of the xylem sap became substantially heavier following irrigation, suggesting that the applied irrigation water was quickly taken up by the plants. TCE concentrations in the xylem sap were diluted by an average of 21% following irrigation; however, dilution was low relative to the increase in sapflux. Our results show that water use by Populus deltoides is variable. Hence, studies addressing phytoremediation effectiveness must account for the relative proportion of surface vs. groundwater uptake.