We have determined the rate of glucose cycling by comparing turnovers of [2-3H]- and [6-3H]glucose under basal conditions and during a glucose infusion. Moreover, the activity of the fructose 6-phosphate cycle was assessed by comparing [3-3H]- and [6-3H]glucose. The study included eight lean subjects with normal glucose tolerance. They participated in two randomly performed investigations. In one experiment [2-3H]- and [6-3H]glucose were given simultaneously, while in the other only [3-3H]glucose was given. The basal rate of glucose cycling was 0.32 +/- 0.08 mg X kg-1 X min-1 or 17% of basal glucose production (P less than 0.005). During glucose infusion the activity of endogenous glucose cycling did not change but since glucose production was suppressed it amounted to 130% of glucose production. The basal fructose 6-phosphate cycle could be detected only in three subjects and was suppressed during glucose infusion. In conclusion, the glucose cycle is active in healthy humans both in basal conditions and during moderate hyperglycemia. In some subjects, the fructose 6-phosphate cycle also appears to be active. Thus it is preferable to use [6-3H]glucose rather than [3-3H]glucose when measuring glucose production and particularly when assessing glucose cycle.