Understanding the strategies that confer resilience on natural woodlands in drought prone environments is important for the conservation of these and similar ecosystems. Our main aim in this 2-year study was to assess traits (sapwood area, sapwood density and leaf area index) that control transpiration in Eucalyptus camaldulensis and E. microcarpa in a natural forest in which topographical variation created surface soils of sandy clay in a depression (clay-zone) and of loamy sand underlain by a dense profile on the terraces (sand-zone). The clay-zone had a wetter profile due to extra water supply through subsurface lateral flow from the adjoining, topographically higher, sand-zone. In the clay-zone, the differences between the two tree species in their hydraulic attributes were large and rates of water use were widely divergent. Rates of transpiration per unit land area (Ec) and canopy conductance of E. camaldulensis that was dominant in the clay-zone were about 50% lower than those for E. microcarpa in the same zone. This was in marked contrast to the behavior of trees growing in the sand-zone where water availability was persistently low and variations in sapwood density, sapwood area and canopy conductance were narrow. This resulted in almost identical rates of water use for the two species in the sand-zone, despite E. microcarpa dominating the stand. Contrary to many previous studies, sapwood density was positively correlated with Ec in these eucalypt species, while the proportion of trunk area assigned to sapwood declined with sapwood density. Consequently in this low rainfall environment, with prolonged dry seasons, dense sapwood safeguards against turgor loss, and possibly xylem embolism, thereby allowing Ec to be sustained under extremely low soil-water availability. We concluded that variation in hydraulic traits is less likely where trees are under persistent water-stress than where the stress is short and relatively mild.We developed single functions for predicting Ec for the two species by integrating their responses to micrometeorological and soil-water conditions.