One of the features which characterizes the brain metabolically is its dependence, under normal circumstances, upon a supply of glucose from the bloodstream as the only major energy source. The theory that glucose enters the brain by carrier-mediated facilitated diffusion is supported by studies in vivo and in vitro. However, quantitative estimates of kinetic parameters, from studies in vivo and in uitro, have shown anomalies (Cooke & Robinson, 1971). The majority of workers studying kinetics in vivo have used anaesthetized animals. This is unsatisfactory because cerebral metabolic rate falls and glucose accumulates during hypothermia and anaesthesia (Brunner et al., 1971 ; Lowry et al., 1964; Strang & Bachelard, 1973). The possibility that these effects may interact with the transport mechanism for glucose justifies a direct comparison in conscious and anaesthetized animals. Male CFW strain mice were starved for 16-20h and placed in an incubator at 33°C. After lOmin their rectal temperatures were recorded and they were injected subcutaneously with 0.2ml Of D-[U-14C]glUCOSe (20pCi/lOOg body wt.) and returned to the incubator. To study glucose influx over a wide range of blood glucose concentrations, groups of animals were injected intraperitoneally with unlabelled glucose (3 or 6g/kg) 1 5min before the labelled glucose. Two parallel series of experiments were performed, in which mice were injected with 70mg of pentobarbitone sodium/kg 30-45min before the labelled glucose, and maintained in the incubator or at room temperature (19-23°C). All animals were killed 5min after the labelled glucose injection. Rectal temperatures were recorded 20s before decapitation, the heads falling directly into Freon. Blood was collected from the severed necks and analysed for glucose and radioactivity. Brains were removed; neutralized acid-soluble extracts were prepared and fractionated into neutral and amino acid fractions. Preliminary experiments had shown that in conscious and anaesthetized mice, the radioactivity in blood increased linearly for 5min after the subcutaneous injection of labelled glucose and that the blood glucose concentration remained relatively constant during the period after labelled glucose administration, even if unlabelled glucose had been administered 15min earlier. The rate of influx of glucose was calculated by a method similar to that of Gaitonde (1965). The variation of glucose influx into the brains of all three groups of mice, with increasing blood glucose concentrations, can be fitted to hyperbolae described by the Michaelis-Menten equation. Derived kinetic data from regression analysis of standard kinetic plots are shown in Table 1. The derived K,,, values for glucose influx in anaesthetized mice agree with the reported values of 6.0-7.2m~ for anaesthetized rats and mice (Buschiazzo eta[., 1970; Growdon et al. (1971); Bachelard et al., 1973) .The K, value for conscious mice of 1 . 6 5 m ~ implies that pentobarbitone anaesthesia lowers the affinity of the carrier for glucose. By substitution of the derived kinetic parameters into the Michaelis-Menten equation, calculations of the rate of glucose influx at any blood glucose concentration can be made. At physiological blood glucose concentrations in conscious mice, the rate of glucose transport is l.l-l.2pmol/min per g, assuming that there is no efflux of glucose from the brain. This is in agreement with the rate of glucose utilization of 1 .Opmol/min per g for the rat brain reported by Gaitonde (1965). Other estimates of cerebral glucose utilization in the rat are in the range 0.3Q-0.6pmol/min per g (McIlwain & Bachelard, 1971; Cremer, 1970). A high range of glucose efflux, even at physiological blood glucose concentrations, would have to be proposed to make the kinetic data compatible with these estimates of cerebral metabolic rate.