These experiments demonstrate that rats can immediately adjust their meal size in response to variations in the caloric density of a novel diet. However, this immediate caloric sensitivity only seems to appear when rats have been adapted to small, calorically insufficient meals. Rats in Experiment 1 were given timed access to unlimited quantities of an oil/water diet during baseline, and they showed no indication of compensating for changes in the caloric density of the oil/water diet during a test meal. Instead, they consumed about the same amount they had consumed during the preceding baseline meal, suggesting that a learned habit of consuming a certain volume of food controlled their meal size. In contrast, rats that were accustomed to receiving only a very small quantity of food for one of their daily meals during baseline immediately responded to the caloric density of an oil/water test diet by consuming a larger meal if the diet was dilute than if it was calorically more concentrated (Experiments 2 and 3). This immediate sensitivity to caloric density occurred whether or not the rats were exposed to the oil/water diet during baseline, suggesting that rats have some way of directly "metering" the caloric density of new foods. Thus, rats' caloric intake during a meal appears to be controlled by two factors: under certain conditions, control is by caloric learning, under other conditions control is by a caloric metering mechanism.