The purposes of this study were to quantify the impact of the duration of infusion and choice of stable isotope of glucose on measures of glucose rate of appearance (glucose R(a)) and to determine whether the differences observed were due to tracer recycling via the glycogen pool (direct pathway) or gluconeogenesis (indirect pathway). Six healthy adult volunteers were studied on four occasions in the postabsorptive state during infusions of [1-(13)C]glucose and [6,6-(2)H(2)]glucose: 2.5-h infusion of both (A), and 2.5-h infusion of one (B) and 14.5-h infusion of the other isotope (C), and 5-h infusion of [6,6-(2)H(2)]glucose and 2.5-h infusion of [1-(13)C]glucose (D). Infusion of both isotopes for 2.5 h resulted in similar glucose R(a) values. When compared with a 14.5-h infusion, the 2.5-h glucose tracer infusion overestimated glucose R(a) by 26-35%. Glucose (13)C recycled via the Cori cycle, resulting in slower decay from the plasma pool and longer half-life of [1-(13)C]glucose compared with [6,6-(2)H(2)]glucose. There was no detectable release of [(13)C]glucose or [(2)H(2)]glucose tracer into the plasma pool after administration of glucagon. These data demonstrate that glucose R(a) varies not as a result of isotope cycling but as a result of differences in duration of isotope infusion regardless of the isotope used. This is most likely due to incomplete isotope and substrate equilibration with the 2.5-h infusion. The potential error was reduced by nearly 80% using a 5-h infusion of [6,6-(2)H(2)]glucose. These studies demonstrate that the duration of isotope infusion has significantly greater impact on quantitation of glucose R(a) than does the selection of isotope.