Over the past several decades, cheatgrass (Bromus tectorum) has been continually expanding in the sagebrush steppe ecosystem. There has been very little research that examines why cheatgrass is able to invade these communities. To determine the effects of herbaceous vegetation removal and fire on available water for cheatgrass invasion, as well as effects of cheatgrass itself, we measured spring soil moisture and vegetation cover on nine sites in Nevada and Utah. Total time of soil water availability was calculated for 1 April through 30 June in 2002 and 2003 on plots that were left intact, sprayed with an herbicide to kill all herbaceous vegetation, burned, or both burned and sprayed in 2001. Soil water was also measured on subplots either left unseeded or seeded to cheatgrass in fall 2001. The number of days that soil moisture was available increased with site elevation and annual precipitation. Time of surface soil (1 to 3 cm) water availability was increased by 12 days in a wetter year (2003) compared to a drier year (2002) and was not significantly affected by disturbance. Time of subsurface soil water availability (13 to 15 and 28 to 30 cm) was increased more by disturbance than seasonal precipitation. Vegetation removal plots had 7 days longer subsurface soil water availability than intact plots the first year after removal. Burned plots had 4 to 6 days longer subsurface soil water availability than unburned plots. Cheatgrass establishment was limited, especially at the Utah sites, and did not significantly affect soil water availability. Soil water resources may be increased by disturbance, higher annual precipitation, or both and thereby facilitate weed invasion.