Use of 13C-labeled glucose for estimating in vivo rates of glucose oxidation faces several difficulties, particularly the accurate determination of the output of 13C in expired air. In an investigation of wholebody glucose metabolism in healthy adult humans, using a continuous intravenous infusion of D-[U-13C]glucose, we found that a precise estimate of the rate of glucose oxidation was difficult to achieve when the study included infusions with unlabeled glucose. Problems arose 1) as a result of the slow rate at which the 13CO2 released by glucose oxidation reaches an equilibrium in expired air CO2 and 2) due to the contribution to 13CO2 output by the natural 13C in the unlabeled glucose that was infused. In a subsequent series of experiments in healthy young adults, we found that the entry of 13CO2 released by the tissues into the bicarbonate pool and into the expired air is relatively slow and a tracer infusion protocol of approximately 6 h is required for determination of glucose oxidation. This applies when metabolic states are changed acutely during the experiment or when unlabeled glucose is infused. However, for resting subjects in the basal postabsorptive state we confirmed that the time required to achieve a steady state in the 13C enrichment of expired air can be shortened significantly by the use of a NaH13CO3 priming dose, even when this dose varies from the ideal.