W e have quantitatively determined gluconeogenesis (GNG) from all precursors, using a novel method employing ²H₂0 to address the question of whether changes in plasma free fatty acids (FFA) affect GNG in healthy, nonobese subjects. In the first study (n = 6), plasma FFA were lowered at 16 to 20 hours with nicotinic acid (NA) and were then allowed to rise at 20 to 24 hours (FFA rebound after administration of NA). FFA decreased from 387 μM at 16 hours to 43 μM at 20 hours, and then rebounded to 1,823 μM at 24 hours. GNG decreased from 58.1% at 16 hours to 38.6% of endogenous glucose production at 20 hours (P < 0.005) and then rebounded to 78.9% at 24 hours (P < 0.05). Conversely, glycogenolysis (GL) increased from 41.9% at 16 hours to 61.4% at 20 hours (P < 0.05), and then decreased to 21.1% at 24 hours (P < 0.05). In the second study (controls; n = 6), volunteers were analyzed between 16 and 24 hours after the last meal. FFA rose from 423 to 681 μM (P < 0.05), and GNG from 50.3% to 61.7% (P < 0.02), whereas GL decreased from 49.7% to 38.3% (P < 0.05). Endogenous glucose production decreased at the same rate in both studies, from 10.7 to 8.6 μmol/kg/min (P < 0.05). In study 3 (n = 6), in which the NA-mediated decrease of plasma FFA was prevented by infusion of lipid and heparin, neither FFA nor GNG changed significantly. In summary, our data suggest that (a) acute changes in plasma FFA produce acute changes in GNG and reciprocal changes in GL; (b) the decrease in EGP between 16 and 24 hours of fasting is due to a fall in GL; and (c) NA has no direct effect on GNG.