A B 5 T RAC T. Bail reform is gaining momentum nationwide. Reformers aspire to untether pretrial detention from wealth (the ability to post money bail) and condition it instead on statistical risk, particularly the risk that a defendant will commit crime if he remains at liberty pending trial. The bail reform movement holds tremendous promise, but it also forces the criminal justice system to confront a difficult question: what statistical risk that a person will commit future crime justifies short-term detention -if any does? What about lesser restraints on liberty, like GPS monitoring? Although the turn to actuarial risk assessment in the pretrial context has raised concern in some quarters, the debate so far has largely ignored this foundational question. One way of thinking about what level of crime risk justifies restraint is to ask whether the answer is different for defendants than for anyone else. It is generally assumed that defendants are a special case, exempt by virtue of pending charges from otherwise applicable protections against preventive interference. This Article challenges that assumption. It argues that, for purposes of restraint for general dangerousness, there is no clear constitutional, moral, or practical basis for distinguishing defendants from non-defendants who are equally dangerous. There is thus no basis to conclude that the risk standard for such restraint should be different for defendants than for anyone else.