Quilpie mesquite (Prosopis velutina Woot.) is an invasive, prickly woody weed that is believed to have been introduced into southwest Queensland in the 1930s. Although it produces palatable pods and provides shade for grazing stock, it develops dense thickets, reducing pasture growth. Field results using triclopyr-based formulations with basal stem application showed inconsistencies that appeared to be related to rainfall patterns. In March 1995, we tested the efficacy of selected herbicides using dieseline as the basal stem herbicide carrier. Herbicide treatments were set out adjacent to the south boundary of Comongin Station Quilpie, Queensland. Plants with root system in a predominately dry condition after spraying were more susceptible to the applied herbicides than plants with greater access to soil moisture, particularly within one month of herbicide application. At the second assessment in November 1995, a distinct pattern of plant degradation occurred because of herbicide-induced plant degeneration, which is linked to water availability. Herbicide efficacy showed significant responses to the effect of water availability (P <0.01), with a significant interaction between water access and herbicide treatment (P <0.01). High soil moisture on the day of application promotes herbicide efficacy, provided follow-up rain did not occur. The higher strength dicamba and triclopyr formulations showed the most effectiveness irrespective of plant height (< 1.5 m or > 1.5 m) or water availability. Smaller plants were more susceptible to applied herbicides. The recovery response of triclopyr + picloram treated plants when close to ponding depressions was possibly because of the buffering capacity provided by the larger biomass of mesquite bordering channels, enabling detoxification of the lower dose fomulations.