Woody encroachment, or the spread of woody species into previously herbaceous-dominated ecosystems, is often attributed to the loss of historic disturbance regimes, such as frequent, low-intensity fire. Once established, woody species can withstand the reintroduction of fire, continue to spread via clonal reproduction, and ultimately lead to the decline of historically dominant herbaceous species. Thus, controlling woody species often requires mechanical and/or chemical management practices in addition to the reintroduction of fire. Management outcomes, however, are inconsistent. One potential reason that is difficult to predict the outcomes of woody management is the concurrent influence of climate change. In many temperate regions, the intensity of summer precipitation events is projected to increase but the frequency decrease, increasing drought-like conditions between events. We designed a full-factorial field experiment to test how different woody management practices and drought impact management outcomes, using physiological variation among plant functional groups to inform the mechanisms underpinning these outcomes. We established eight 32-m2 plots randomly placed within patches of gray dogwood (Cornus racemosa) – a prevalent encroaching species of high management concern – at two unplowed remnant tallgrass prairies in southern Wisconsin, one dry and one mesic. We treated each plot with one of four woody management types (mechanical, chemical, mechanical and chemical, or unmanaged) and either ambient or reduced precipitation (i.e., drought) using rainout shelters. We collected data on management outcomes in nested 10-m2 subplots by estimating the percent cover of gray dogwood. We also collected physiology data – including pre-dawn (ΨPD) and mid-day (ΨMD) leaf water potential and mid-day photosynthetic rate (Anet) and stomal conductance (gs) – on gray dogwood and the most dominant graminoids (Sorghastrum nutans, Carex stipata) and forbs (Helianthus pauciflorus, H. grosseserratus) at each site. We found a clear impact of management on gray dogwood, as percent cover significantly declined in both chemical treatment types (chemical and mechanical and chemical). However, management outcomes were much more variable under drought conditions, and mechanical and unmanaged treatments trended toward greater gray dogwood cover under drought. We also found significantly lower ΨPD and ΨMD across species under drought than ambient conditions, with greater differences between drought and ambient conditions in ΨPD than ΨMD. However, the drought effects were greater for the dominant herbs than for gray dogwood. Graminoids and forbs showed large fluctuations in water potential, whereas gray dogwood showed less response to daily and especially treatment-caused water fluctuations. With this, gas exchange did not vary under drought conditions in gray dogwood, while graminoids and forbs neared zero Anet. These findings provide important insights into the differences in drought stress response among an encroaching woody species and dominant herbaceous species, suggesting that encroachment may continue to worsen under drought conditions where woody species are already well established, even with management.