Although negative hepatic glucose production rates are physiologically impossible, they have been observed when hepatic glucose production is measured with the tracer-dilution technique during the hyperinsulinemic, euglycemic glucose clamp. Because hepatic glucose production is determined from the difference between tracer-derived glucose disposal and the known exogenous glucose infusion rate, the negative values for hepatic glucose production must result from an underestimation of glucose disposal by the tracer technique. In the current investigation, tracer-derived glucose disposal was measured in 25 subjects undergoing hyperinsulinemic, euglycemic clamps. Glucose disposal was measured with both radioactive and stable isotopes that utilize different methodologies, to determine whether discriminant metabolism of the isotopes versus methodological error leads to underestimation of tracer-derived glucose disposal. Both the radioactive and stable methodologies underestimated the exogenous glucose infusion rate during the hyperinsulinemic euglycemic clamp by 27 and 17%, respectively. Mean hepatic glucose production was -2.1 +/- 0.2 and -1.3 +/- 0.2 mg X kg-1 X min-1 as determined by the radioactive and stable isotope methodologies, respectively. Methodological error was an unlikely cause of this underestimation because it occurred with two different methodologies. The most likely explanation for underestimated rates of glucose disposal determined by the two types of isotope methodologies is discrepant metabolism of glucose tracers in comparison with unlabeled glucose.