The paper introduces the concept of adjustment utility, that is, reference-dependent utility from expectations. It offers an explanation for observed preferences that cannot be explained with existing models, and yields new predictions for individual decision making. The model gives a simple explanation for, e.g., why people are reluctant to change their plans even when these turn out to be unexpectedly costly; people’s aversion towards positive but false information, which cannot be explained with previous models; and the increasing acceptance of risks when people get used to them.