Critical illness is associated with a marked increase in metabolic rate and progressive wasting, despite aggressive nutritional support. The metabolic events which are responsible for these phenomena are unclear, but are characterised by marked impairment of the anabolic effects of insulin on glucose metabolism and excessive activation of the sympathetic nervous system. It has been suggested that critical illness may be associated with impaired carbohydrate oxidation and a marked increase in the loss of heat energy associated with glucose administration (glucose-induced thermogenesis). This situation may result in impaired efficiency of nutrient assimilation. Studies employing combinations of nutrient infusions both at clinically-relevant rates and in association with euglycaemic hyperinsulinaemia have, however, demonstrated that nutrient-induced thermogenesis is unaffected in critical illness in human subjects, and that defective glucose utilization occurs as a consequence of impaired insulin-mediated glucose storage rather than oxidation. Although the cellular and molecular mechanisms underlying these changes are controversial, the recent validation of a human model of insulin resistance in critical illness should provide a means of studying this response in future, and allow the identification of therapeutic targets. This information should increase the efficacy of nutritional support in some of our most seriously-ill patients.