It is usually assumed that absorbed nutrients have effects which contribute to the control of food intake. Indeed, we have suggested that the rate of flow of energy to working tissues is the basic controlling factor in appetite and satiety. The chief absorbed energy substrate in the rat on a laboratory diet is glucose: yet glucose administration does not always produce a decrease in energy intake which compensates for all the glucose energy administered. Such negative findings may be artefacts of failure to mimic the absorption of glucose during normal satiety; sometimes, for instance, as has been pointed out, undetectably low doses, unphysiological concentrations and volumes, and starved rats that are not easily satiated have been used. Furthermore, it may be necessary to produce the normal context for nutrient utilisation, such as adequate secretion of insulin. For example, the reduction in food intake during continuous intravenous infusion of glucose is at most 0.6 kcal per kcal infused: coinfusion of insulin corrects this incomplete compensation. We report here that the context of the initial absorption of a meal on familiar diet permits full compensation for infused glucose by a decrease in meal size. In contrast, when an unfamiliar cue is added to the diet, conditioning of appetite by the infusion is sufficient to overwhelm satiety, disrupting compensation and causing faster weight gain.