Trees exposed to elevated CO2 partial pressure ([CO2]) generally show increased rates of photosynthesis and growth, but effects on leaf respiration are more variable. The causes of this variable response are unresolved. We grew 12-year-old sweetgum trees (Liquidambar styraciflua L.) in a Free-Air CO2 Enrichment (FACE) facility in ambient [CO2] (37/44 Pa daytime/nighttime) and elevated [CO2] (57/65 Pa daytime/nighttime) in native soil at Oak Ridge National Environmental Research Park. Nighttime respiration (R(N)) was measured on leaves in the upper and lower canopy in the second (1999) and third (2000) growing seasons of CO2 fumigation. Leaf respiration in the light (R(L)) was estimated by the technique of Brooks and Farquhar (1985) in the upper canopy during the third growing season. There were no significant short-term effects of elevated [CO2] on R(N) or long-term effects on R(N) or R(L), when expressed on an area, mass or nitrogen (N) basis. Upper-canopy leaves had 54% higher R(N) (area basis) than lower-canopy leaves, but this relationship was unaffected by CO2 growth treatment. In August 2000, R(L) was about 40% of R(N) in the upper canopy. Elevated [CO(2)] significantly increased the number of leaf mitochondria (62%), leaf mass per unit area (LMA; 9%), and leaf starch (31%) compared with leaves in ambient [CO(2)]. Upper-canopy leaves had a significantly higher number of mitochondria (73%), N (53%), LMA (38%), sugar (117%) and starch (23%) than lower-canopy leaves. Growth in elevated [CO2] did not affect the relationships (i.e., intercept and slope) between R(N) and the measured leaf characteristics. Although no factor explained more than 45% of the variation in R(N), leaf N and LMA were the best predictors for R(N). Therefore, the response of RN to CO2 treatment and canopy position was largely dependent on the magnitude of the effect of elevated [CO2] or canopy position on these characteristics. Because elevated [CO2] had little or no effect on N or LMA, there was no effect on R(N). Canopy position had large effects on these leaf characteristics, however, such that upper-canopy leaves exhibited higher R(N) than lower-canopy leaves. We conclude that elevated [CO2] does not directly impact leaf respiration in sweetgum and that barring changes in leaf nitrogen or leaf chemical composition, long-term effects of elevated [CO2] on respiration in this species will be minimal.