This Article continues our project of applying new findings in the behavioral psychology of human happiness to some of the most deeply analyzed questions in law. When a state decides how to punish criminal offenders, at least one important consideration is the amount of harm any given punishment is likely to inflict. It would be undesirable, for example, to impose greater harm on those who commit less serious crimes or to impose harm that rises to the level of cruelty. Our penal system fits punishments to crimes primarily by adjusting the size of monetary fines and the length of prison terms, but new findings about human adaptability unsettle the assumptions upon which the system rests. Specifically, people adapt well to negative changes in wealth and even to many features of prison life, whereas they adapt poorly to typical conditions of postprison life such as unemployment, disease, and loss of social ties. As a result, adjusting the size of a fine or the length of a prison term will not change in a linear fashion the amount of harm imposed. A large difference in the size of two punishments will not necessarily result in a large (or, in the case of some fines, any) difference in the harm felt by the respective offenders. This result is relevant to both retributive and utilitarian punishment theories, as well as to the practice of criminal justice. New approaches to punishment are necessary to achieve proportionality in light of the challenges posed by adaptation.