The relationship between gluconeogenic precursor supply and glucose production has been investigated in 14-h and 86-h fasted humans. In protocols 1 and 2 [6,6-2H]glucose and [15N2]urea were infused to measure glucose and urea production rates (Ra) in response to infusions of glycerol and alanine. In protocol 3 first [15N]alanine, [3-13C]lactate, and [6,6-2H]glucose were infused before and during administration of dichloroacetate (DCA) to determine the response of glucose Ra to decreased fluxes of pyruvate, alanine, and lactate, then alanine was infused with DCA and glucose Ra measured. After a 14-h fast, neither alanine nor glycerol increased glucose Ra. Basal glucose Ra decreased by one-third after 86 h of fasting, yet glycerol and alanine infusions had no effect on glucose Ra. Glycerol always reduced urea Ra (P less than 0.05), suggesting that glycerol competitively inhibited gluconeogenesis from amino acids. DCA decreased the fluxes of pyruvate, alanine (P less than 0.01), and glucose Ra (P less than 0.01), which was prevented by alanine infusion. These findings suggest that 1) the reduction in glucose Ra after an 86-h fast is not because of a shortage of gluconeogenic substrate; 2) nonetheless, the importance of precursor supply to maintain basal glucose Ra is confirmed by the response to DCA; 3) an excess of one gluconeogenic substrate inhibits gluconeogenesis from others.