Desiccation of the vadose zone has the potential to reduce the flux of contaminants to underlying groundwater by removing moisture and decreasing the water relative permeability of the desiccated zone. However, data to evaluate implementation of desiccation are needed before desiccation can be considered as a potential remedy. Implementation of desiccation was field tested by injecting dry nitrogen gas to a target treatment zone and monitoring the spatial and temporal progress of the drying process. Aqueous waste discharges to disposal cribs approximately 50 yr ago distributed water and contaminants, primarily technetium-99 and nitrate, within the 100-m deep vadose zone at the test site. The test was conducted adjacent to one of the former disposal cribs in a contaminated portion of the vadose zone dominated by fine sands with lenses of loamy sand. Desiccation removed over 18,000 kg of water from the test zone within the 151-d active desiccation period and reduced volumetric moisture content over 1300 m3 of soil with values lower than 0.01 (m3 m−3) in 68 m3. The lateral and vertical distribution of drying from the injection well was influenced by the subsurface heterogeneity with initial drying in higher permeability zones. However, over time, desiccation also occurred in the initially wetter, lower permeability lenses.