D espite ample evidence that the kidney can both produce and use appreciable amounts of glucose, the human kidney is generally regarded as playing a minor role in glucose homeostasis. This view is based on measurements of arteriorenal vein glucose concentrations indicating little or no net release of glucose. However, inferences from net balance measurements do not take into consideration the simultaneous release and uptake of glucose by the kidney. Therefore, to assess the contribution of release and uptake of glucose by the human kidney to overall entry and removal of plasma glucose, we used a combination of balance and isotope techniques to measure renal glucose net balance, fractional extraction, uptake and release as well as overall plasma glucose appearance and disposal in 10 normal volunteers under basal postabsorptive conditions and during a 3-h epinephrine infusion. In the basal postabsorptive state, there was small but significant net output of glucose by the kidney (66 +/- 22 mumol.min-1, P = 0.016). However, since renal glucose fractional extraction averaged 2.9 +/- 0.3%, there was considerable renal glucose uptake (2.3 +/- 0.2 mumol.kg-1.min-1) which accounted for 20.2 +/- 1.7% of systemic glucose disposal (11.4 +/- 0.5 mumol.kg-1.min-1). Renal glucose release (3.2 +/- 0.2 mumol.kg-1.min-1) accounted for 27.8 +/- 2.1% of systemic glucose appearance (11.4 +/- 0.5 mumol.kg-1.min-1). Epinephrine infusion, which increased plasma epinephrine to levels observed during hypoglycemia (3722 +/- 453 pmol/liter) increased renal glucose release nearly twofold (5.2 +/- 0.5 vs 2.8 +/- 0.1 mol.kg-1.min-1, P = 0.01) so that at the end of the infusion, renal glucose release accounted for 40.3 +/- 5.5% of systemic glucose appearance and essentially all of the increase in systemic glucose appearance. These observations suggest an important role for the human kidney in glucose homeostasis.