When newborn dogs 2–14 days of age, anaesthetized with Nembutal, were cooled to 30°C body core temperature, their rate of CO2 production (QCO2) decreased as predicted by van’t Hoff’s law, but the rates of hepatic glucose production and overall glucose utilization were decreased to a lesser degree, indicating an increase in the percentage contribution of plasma glucose to the overall energy turnover. In different litters of pups cooled to 32°C, QCO2 and the rate of glucose production and utilization were decreased as predicted by van’t Hoff’s law. The metabolic clearance rate of glucose was decreased by cooling in all pups. These changes were wholly or partially reversible upon rewarming. Because of the approximate parallelity in the changes in the rates of production and utilization of glucose, no consistent change in blood glucose was observed during cooling. When non-anaesthetized pups less than 6 days old were exposed to cold until their core temperature was decreased to 32 °C: QCO2 increased transiently during cooling. A small and transient elevation in plasma glucose concentration was observed due to a small increase in hepatic glucose production. Older pups fell into two categories: those which were able to maintain their body temperature by increasing their metabolic rate did not react with any change in their rates of glucose production and utilization. On the other hand pups which, during cooling, failed to raise their metabolic rate became hyperglycaemic because of an increased rate of glucose production and a decrease in glucose clearance. In conclusion: in anaesthetized pups QCO2 and the rates of hepatic glucose production and utilization follow approximately van’t Hoff’s law, indicating the absence of significant regulatory responses. Non-anaesthetized pups, if they are able to maintain their body temperature during cooling by an increase of their metabolic rate, show no significant changes in glucose production and utilization. Hyperglycaemia developed only in those pups which failed to increase their metabolic rate and consequently became severely hypothermic. In no case did hypothermia lead to hypoglycaemia.