The effects of elevated carbon dioxide (CO2) concentration on plant water use are best evaluated on plants grown under field conditions and with measurement techniques that do not disturb the natural function of the plant. Heat balance sap flow gauges were used on individual main stems of wheat (Triticum aestivum L. cv Yecora rojo) grown under normal ambient conditions (control) and in a free-air CO2 enrichment (FACE) system in Arizona with either high (control + high H2O = CW; FACE + high H2O = FW) or low (control + low H2O = CD; FACE + low H2O = FD) irrigation regimens. Over a 30d period (stem elongation to anthesis), combinations of treatments were monitored with,10–40 gauges per treatment. The effects of increased CO2 on tiller water use were inconsistent in both the diurnal patterns of sap flow and the statistical analyses of daily sap flow (Ftot). Initial results suggested that the reductions in Ftot, from CO2 enrichment were small (,0–10%) in relation to the H2O treatment effect (,20–30%). For a 3d period, Ftot of FW was,19–26% less than that of CW (P = 0.10). Examination of the different sources of variation in the study revealed that the location of gauges within the experimental plots influenced the variance of the sap flow measurements. This variation was probably related to positional variation in subsurface drip lines used to irrigate plots. A sampling design was proposed for use of sap flow gauges in FACE systems with subsurface irrigation that takes into account the main treatment effects of CO2 enrichment and the other sources of variation identified in this study. Despite the small and often statistically non-significant differences in Ftot between the CW and FW treatments, cumulative water use of the FW treatment at the end of the first three test periods ranged from 7 to 23% lower than that of the CW treatment. Differences in sap flow between FW and CW compared well with treatment differences in evapotranspiration. The results of the study, based on the first reported sap flow measurements of wheat, suggest that irrigation requirements for wheat production, in the present climatic regimen of the south-western US, may be predicted to decrease slightly because of increasing atmospheric CO2.