Effects of landscape design and land use history on gas exchange parameters were evaluated for woody plants in a factorial site matrix of formerly desert or agricultural land uses and xeric or mesic residential landscape designs within the metropolitan area of Phoenix, Ariz. Remnant Sonoran Desert sites and an alfalfa agricultural field functioned as controls. Residential landscapes and the alfalfa field were irrigated regularly. Monthly instantaneous measurements of maximum leaf and stem carbon assimilation (A), conductance (gs), and transpiration (E) were made within three replicates of each site type during 1998 and 1999. Measurements were repeated monthly on three woody plant life forms: trees, shrubs, and ground covers. Assimilation fluxes were not related to former land use, but were lower for plants in xeric compared with those in mesic landscapes. Transpiration fluxes were higher for plants in formerly agricultural sites than in formerly desert sites, and were lower in xeric than in mesic landscape design. Compared with plants in residential landscapes, A and E fluxes were generally higher for plants in the agricultural control sites and were lower for plants at the desert control sites. Plant instantaneous transpiration efficiency (ITE = A/E) was higher in formerly agricultural sites than in formerly desert sites but was not affected by landscape design. Patterns of A, gs, and shoot temperature at irrigated sites suggest that maximum plant carbon assimilation was not limited by shoot conductance but was more responsive to shoot temperature. Similarities in patterns of ITE between plants in the different landscape design types suggest that xeric and mesic landscape plants do not differ in terms of water use efficiency.