The effects of sodium dichloroacetate (DCA) on carbohydrate metabolism in conscious, 48-h-fasted dogs were examined using the hepatic A-V difference technique and a double isotope infusion technique (3H-glu-cose to measure glucose production and 14C-alanine to assess gluconeogenesis). DCA infusion (0.4 mg/kg-min) resulted in an 82 ± 1% fall in the arterial plasma alanine level and a 53 ± 8% fall in the arterial whole blood lactate level. Hepatic uptake of alanine and lactate fell 67 ± 5% and 59 ± 15%, respectively, although the fractional extraction of these intermediates was not altered. DCA decreased the conversion of circulating alanine and lactate to glucose but by only 41 ± 7%, suggesting that a slight increase in the efficiency of the intrahepatic gluconeogenic process took place. This may be explained by the decrease in the plasma insulin level (39 ± 9%) that occurred in the presence of an unchanged plasma glucagon concentration. Despite the substantial fall in the levels of gluconeogenic precursors in blood and the considerable decrease in their rate of conversion to glucose, the overall rates of glucose production and the blood glucose concentration were not altered by DCA. These data indicate that the alanine and lactate supplied by the periphery after a 48 h fast in the dog are not essential for the acute maintenance of glucose production or euglycemia. They suggest, further, that a compensatory increase in glucose production can occur by drawing on an alternate intrahepatic carbon source, the nature of which and signal for which remain unclear.