The δ ¹³C of slash pine foliage dark-respired CO₂ (δ ¹³C_F) was measured at temperatures ranging from 6.6° to 30°C. The δ ¹³C_F was ¹³C-enriched initially after dark by as much as 9‰ and declined toward an asymptotic value after a few hours in the dark. Maximum enrichment in δ ¹³C_F occurred at the lowest temperatures and the decline in δ ¹³C_F toward the asymptotic value took longer at lower temperatures. The effect of this temporal variation in foliage enriched CO₂ on determination of the δ ¹³CO₂ of ecosystem respiration (δ ¹³C_R) was investigated using two approaches by measuring the nighttime CO₂ concentration and δ ¹³CO₂ ratio in a 30-m-tall pine forest in Florida, United States. The first approach (time series) consisted of making measurements at a single height at midcanopy from 1 hour after sunset until 0200 local time. The second approach (profile) consisted of making measurements at multiple heights from ground level to the top of the canopy synoptically prior to dawn. For each method a Keeling plot was used to estimate the δ ¹³C_R. The δ ¹³C_R values determined with the time series were ¹³C-enriched relative to those determined from profiles on average by 1.0‰ and by as much as 3.1‰. A mass balance was used to determine the impact of the nonconstancy of δC_F on the overall Keeling plot intercepts. For the months examined, results indicate that (1) the enrichment in the Keeling intercepts determined from time series compared to those determined from profiles was consistent with the addition of nighttime foliage respired ¹³C-enriched CO₂ and (2) Keeling plots determined from midcanopy heights appear linear despite the addition of temporally variable source δ ¹³CO₂ (range: 4‰). A common approach that specifies the time (predawn) and spatial distribution (multiple heights) for CO₂ collection is suggested to improve intercomparisons of δ ¹³C_R from different ecosystems.