In many places along the lower Colorado River, saltcedar (Tamarix spp) has replaced the native shrubs and trees, including arrowweed, mesquite, cottonwood and willows. Some have advocated that by removing saltcedar, we could save water and create environments more favourable to these native species. To test these assumptions we compared sap flux measurements of water used by native species in contrast to saltcedar, and compared soil salinity, ground water depth and soil moisture across a gradient of 200–1500 m from the river's edge on a floodplain terrace at Cibola National Wildlife Refuge (CNWR). We found that the fraction of land covered (f_c) with vegetation in 2005–2007 was similar to that occupied by native vegetation in 1938 using satellite-derived estimates and reprocessed aerial photographs scaled to comparable spatial resolutions (3–4 m). We converted f_c to estimates of leaf area index (LAI) through point sampling and destructive analyses (r² = 0·82). Saltcedar LAI averaged 2·54 with an f_c of 0·80, and reached a maximum of 3·7 with an f_c of 0·95. The ranges in f_c and LAI are similar to those reported for native vegetation elsewhere and from the 1938 photographs over the study site. On-site measurements of water use and soil and aquifer properties confirmed that although saltcedar grows in areas where salinity has increased much better than native shrubs and trees, rates of transpiration are similar. Annual water use over CNWR was about 1·15 m year⁻¹.