The "harm principle," which suggests that the State should criminalize conduct only if doing so is necessary to prevent harm to others, has long been at the center of debates about the legal enforcement of morality through the law. This Essay presents a series of original experiments that investigate the psychological "plasticity" of harm in the context of criminalization-i.e., the extent to which people may recruit harm to achieve desired outcomes within the terms of a given legal constraint. In the first study, we identified various scenarios that induced defiance of the harm principle, such that the participants wanted to criminalize the conduct even if they did not think that it caused harm to others. We then used these scenarios in a second study to investigate how people would respond when they were told that the law requires a finding of harm in order to impose a criminal penalty. As predicted, the respondents who were presented with the necessity-of-harm constraint continued to criminalize at the same rate, but they imputed harm to the conduct where harm was not previously reported. The third study demonstrated the directionality and universality of the harm plasticity effect by showing that a legally extrinsic ideological factor exacerbated the recruiting of harm among people on both sides of a controversial issue (abortion), without their recognition. The final study provided evidence for the nonintentional nature of this effect and pointed toward a pathway to potential remedies. These experiments reveal motivated, outcome-driven perceptions of harm that are at odds with the objective, perception-driven outcomes toward which the legal system strives. We discuss the implications of our findings for the criminal regulation of morality, constitutional law, and theories of punishment and moral reasoning. Furthermore, we consider the application of our results to legal decision making by lay people and professional adjudicators in the context of two high- profile Ninth Circuit cases. We conclude with a discussion of how to potentially reduce motivated cognition in legal judgments.