1. During infusion of [5-15N]glutamine in patients with gastrointestinal cancer we unexpectedly observed a gradual decrease in time of the appearance rate (Ra) of glutamine in plasma. Here we investigate whether the failure to achieve a plateau isotopic enrichment in plasma is, among other factors, due to incomplete equilibration of the glutamine tracer with the large intramuscular free glutamine pool.2. Plasma and intramuscular glutamine enrichment were measured during 6-11 h infusions of L-[5-15N]glutamine and L-[1-13C]glutamine in post-absorptive patients admitted to hospital for elective abdominal surgery. L-[1-13C]Leucine and L-[ring-2H5]phenylalanine were infused to measure the proportion of glutamine appearing in plasma directly due to its release from protein.3. The glutamine tracer entered muscle, but the rise in intramuscular glutamine enrichment was small, presumably as a result of the enormous size of the intramuscular glutamine pool and the limited speed of entry of glutamine into muscle. In each patient the intramuscular glutamine enrichment was lower than that in plasma (P<0.001), and both increased with tracer infusion time (P<0.001), indicating incomplete equilibration of the glutamine tracer.4.A comparison of the results obtained by the two glutamine tracers indicated that recycling of the nitrogen label contributed to about 15% of the decrease in Ra.5. There was a gradual reduction in the glutamine release from proteolysis, which contributed to 16-21% of the decline in Ra.6. We conclude that slow equilibration of the glutamine tracer with the large muscle glutamine pool significantly contributes to the absence of isotopic steady state. Consequently, the appearance rate of glutamine in plasma measured during short tracer infusion periods (hours) considerably overestimates the whole-body glutamine flux.