A strong correlation was previously observed between carbon isotope discrimination (⌬ 13 C) of phloem sap sugars and phloem sap sugar concentration in the phloem-bleeding tree Eucalyptus globulus Labill. (J. Pate, E. Shedley, D. Arthur, M. Adams [1998] Oecologia 117: 312-322). We hypothesized that correspondence between these two parameters results from covarying responses to plant water potential. We expected ⌬ 13 C to decrease with decreasing plant water potential and phloem sap sugar concentration to increase, thereby maintaining turgor within sieve tubes. The hypothesis was tested with analyses of E. globulus trees growing on opposite ends of a rainfall gradient in southwestern Australia. The ⌬ 13 C of phloem sap sugars was closely related to phloem sap sugar concentration (r ϭ Ϫ0.90, P Ͻ 0.0001, n ϭ 40). As predicted, daytime shoot water potential was positively related to ⌬ 13 C (r ϭ 0.70, P Ͻ 0.0001, n ϭ 40) and negatively related to phloem sap sugar concentration (r ϭ Ϫ0.86, P Ͻ 0.0001, n ϭ 40). Additional measurements showed a strong correspondence between predawn shoot water potential and phloem sap sugar concentration measured at midday (r ϭ Ϫ0.87, P Ͻ 0.0001, n ϭ 30). The ⌬ 13 C of phloem sap sugars collected from the stem agreed well with that predicted from instantaneous measurements of the ratio of intercellular to ambient carbon dioxide concentrations on subtending donor leaves. In accordance, instantaneous ratio of intercellular to ambient carbon dioxide concentrations correlated negatively with phloem sap sugar concentration (r ϭ Ϫ0.91, P Ͻ 0.0001, n ϭ 27). Oxygen isotope enrichment (⌬ 18 O) in phloem sap sugars also varied with phloem sap sugar concentration (r ϭ 0.91, P Ͻ 0.0001, n ϭ 39), consistent with predictions from a theoretical model of ⌬ 18 O. We conclude that drought induces correlated variation in the concentration of phloem sap sugars and their isotopic composition in E. globulus.Measurement of stable carbon and oxygen isotope ratios in plant material provides a valuable tool for studying the performance of terrestrial plants. For example, the strong correlation between discrimination against 13 C (⌬ 13 C) and the ratio of intercellular to ambient carbon dioxide concentrations (c i /c a ) has been relied upon extensively to assess plant water use efficiency under a variety of experimental and natural conditions (for review, see Farquhar et al., 1989a;Ehleringer, 1993;Brugnoli and Farquhar, 2000). Farquhar et al. (1982) derived an expression relating ⌬ 13 C to c i /c a for C 3 photosynthesis such that:where a is the fractionation caused by gaseous diffusion (4.4‰), and b is the effective fractionation caused by carboxylating enzymes (approximately 27‰). The ⌬ 13 C is defined with respect to atmospheric CO 2 as ⌬ 13 C ϭ R a /R p Ϫ 1, where R a is 13 C/ 12 C of atmospheric CO 2 and R p is 13 C/ 12 C of plant material. Equation 1 suggests that ⌬ 13 C decreases linearly as c i /c a decreases. Because c i /c a represents a balance between the supply of CO 2 via stomata and the photosynt...