Two open-circuit indirect-calorimetry ventilated hoods were built for measuring the energy exchanges of cattle nourished wholly be continuous intragastric infusions of mixtures of volatile fatty acids and casein. In Expt. 1 the energy costs of eating and ruminating oat husks were measured. The energy costs of eating and rumination were estimated to be 11.4 J/kg live weight and 9.3 J/kg live weight per minute spent eating and ruminating respectively. The energy cost of standing above that of lying was calculated to be 6.6 kJ/kg live weight per day. In the subsequent four experiments, the effect of abomasal infusions of various amount of glucose on the heat production, nitrogen metabolism and blood metabolites of Friesian steers was investigated. In Expt. 2, the abomasal infusion of 350 g of glucose/d after a five day fast, decreased the urinary nitrogen excretion to the endogenous nitrogen loss values. In Expt. 3 the infusion of glucose at low rates (16, 25 or 50 kJ/kg W^0.75/d) had virtually no nitrogen-sparing effect. In Expts. 4 and 5 the response in heat production to abomasal infusions of glucose differed between steers. Two steers had consistent heat decrements whereas one steer showed mostly heat increments. Fasting urinary nitrogen excretion was gradually reduced by glucose infusion, demonstrating the nitrogen-sparing effect of glucose. Plasma insulin was decreased during the fast in Expt. 4, but it was not affected by glucose infusion in both Expts. 4 and 5, suggesting that nitrogen-sparing is not mediated by insulin. Plasma free fatty acids and β-hydroxybutyrate concentrations were increased several-fold during the fast and were decreased by the infusion of glucose. It is concluded that the increased loss of nitrogen in fasting cattle is unrelated to a specific requirement for glucose precursors during the fast.