The ecohydrological separation hypothesis states that transpiration through plants and drainage to streams and groundwater are sourced from separate soil water pools, which possess distinct isotopic signatures. Evidence for ecohydrological separation has relied on the globally ubiquitous observation that plant water and draining water are isotopically distinct. We evaluated the ecohydrological separation hypothesis in the Dry Creek Experimental Watershed in the semiarid, snow-dominated landscape of southwest Idaho, USA. We found that plant water is indeed isotopically distinct from streams and groundwater. However, we were unable to track those waters to subsurface soil waters, nor were we able to relate soil water mobility to isotopic composition. Soil waters of different mobility can be isotopically similar, and isotopic distinction in soil water can occur for reasons not related to mobility. We suggest that isotopic distinction between root-absorbed and draining waters may not be an appropriate indicator of ecohydrological separation of soil waters, and that hydrologic explanations for such isotopic distinction may not be sufficient.