Abstract Gluconeogenesis from lactate, pyruvate, fructose, glycerol, amino acids, and other compounds was examined in the isolated rat liver perfused with buffer containing blood cells and serum albumin. The following observations were made. The condition of the liver remained good during 2 hours of perfusion at 37° as judged by its gross appearance, oxygen uptake, bile production, water content, and linear rates of glucose production from various substrates. The highest rate of glucose production (about 120 µmoles per g of liver per hour) was observed with saturating concentrations of fructose or dihydroxyacetone. This rate was double that with a saturating level of l-lactate or pyruvate, suggesting that the steps limiting gluconeogenesis from lactate or pyruvate lie before the formation of triose phosphate. The concentration of lactate for half-maximum gluconeogenesis was about 2 mm and that of pyruvate, about 1 mm. The same maximum rate of gluconeogenesis was observed with both substrates. Glycerol, d-glyceraldehyde, l-alanine, or l-serine at high concentrations supported rates of gluconeogenesis about double that found without added substrate (or approximately one-half the maximum rate with l-lactate). d-Lactate, other amino acids, and Krebs cycle intermediates were poor precursors when added singly. With a mixture of amino acids at plasma concentrations, basal glucose production was increased 3-fold and with 1 mm lactate it was doubled. The mean time for conversion of lactate to glucose was 75 sec. The rate of glucose utilization by the liver was low compared with gluconeogenesis. High levels of glucose did not suppress gluconeogenesis from saturating concentrations of lactate. Glucose was the major product of lactate, pyruvate, and fructose metabolism. Lactate production with fructose was greater than that with glucose. With pyruvate-2-14C, there was a 3-fold greater incorporation of isotope into glucose plus glycogen than into CO2. With pyruvate-1-14C, labeling of hexose was much less than that of CO2 and was 38% of that found with pyruvate-2-14C. A mathematical treatment of the isotopic data was developed. With certain assumptions, it was computed that pyruvate was converted to oxalacetate at twice the rate at which it was converted to acetyl coenzyme A, and that most of the acetyl-CoA formed from pyruvate entered the Krebs cycle. It was also deduced that the rate of conversion of oxalacetate to phosphopyruvate was 3-fold greater than the rate of conversion to citrate. There was no correlation between the rates of gluconeogenesis and ketogenesis. Lactate increased glucose production 4-fold but decreased ketogenesis by 70%. Physiological levels of oleate stimulated ketogenesis but not gluconeogenesis in livers from fasted or fed rats. In fed rats, octanoate increased ketogenesis 3-fold but did not affect gluconeogenesis.