The hemlock woolly adelgid (HWA, Adelges tsugae Annand) is causing widespread decline and mortality of eastern hemlock trees (Tsuga canadensis (L.) Carr.). Stem injection of insecticide is widely used as a control measure, but its effectiveness depends on individual tree hydraulic characteristics. Recent work has shown that eastern hemlock daily water use is exponentially related to tree diameter, with smaller-diameter trees using significantly less water than largerdiameter trees. In this study we modeled daily water use for 20 eastern hemlock trees across a range of diameters. Based on expected daily water use and, thus, potential xylem transport of insecticide, we applied a dosage estimated to achieve a lethal and uniform xylem sap concentration of imidacloprid to half the trees (xylem transport treatment), and treated the remaining half based on the manufacturerrecommended dosage (MFR treatment), which is a linear function of tree diameter. At 4 and 56 weeks after treatment, we assessed all trees for the presence or absence of new shoot growth and live HWA population density. We found that both treatment dosages significantly reduced live HWA populations; however, 4 weeks after treatment, live HWA population density was 32 percent lower (LSMEANS) on xylem transport treatment trees compared with MFR treatment trees (treatment by time interaction, P = 0.006). Both treatment dosages also significantly increased the proportion of new shoot growth; however, over time xylem transport treatment trees had a significantly greater increase in new shoot production compared with MFR treatment trees. We conclude that dosages based on a xylem transport model not only significantly improved tree health more than the trees receiving the current recommended dosage, but also caused greater reductions in the live HWA population. Improving treatment dosages would reduce treatment cost, nontargeted effects, and would allow land managers to treat a greater number of trees.