It has been suggested that regulation of glucose-6-phosphatase by insulin plays a role in the suppression of hepatic glucose production during feeding. We used hepatic glucose production (measured with the D-[3-3H]glucose infusion method) as an indicator of substrate flux through glucose-6-phosphatase in vivo. Compared with saline controls, insulin (7 mU.min-1 x kg-1, euglycemic clamp) suppressed hepatic glucose production virtually completely in both fasted (32.4 +/- 2.4 vs. -6.1 +/- 14 mumol.min-1 x kg-1) and fed (64.6 +/- 6.4 vs. 5.5 +/- 5.2 mumol.min-1 x kg-1) rats. Whereas hepatic glucose production was totally suppressed, [glucose-6-phosphate] in liver cytosol declined by only 27 and 35% in fasted and fed rats, respectively. Addition of hyperglycemia (10 mM) to the insulin infusion likewise fully suppressed hepatic glucose production (26.9 +/- 1.4 vs. -9 +/- 10 mumol.min-1 x kg-1 and 80.8 +/- 10.1 vs. -3.6 +/- 12.6 mumol.min-1 x kg-1 in fasted and fed rats, respectively), but [glucose-6-phosphate] again declined only modestly (21 and 27% in fasted and fed rats, respectively). This disproportionate suppression of hepatic glucose production could not be explained by cooperative substrate effects inasmuch as microsomal glucose-6-phosphatase isolated from saline- and insulin-treated rats followed Michaelis-Menten kinetics (Hill coefficient approximated 1). Acute insulin treatment of fasted rats in vivo did not reproducibly inhibit glucose-6-phosphatase activity assayed subsequently in isolated microsomes incubated in the absence of insulin.(ABSTRACT TRUNCATED AT 250 WORDS)