Abstract: ...In this Article, I analyze the theoretically optimal use of nonmonetary sanctions as a deterrent, and I then examine their actual domain of use and the major principles and doctrines of the criminal law. ... Otherwise, there would be an opportunity to lower social costs while continuing to deter the same number of parties by simultaneously reducing the nonmonetary sanction and increasing the monetary sanction. ... The initial three factors are important because they bear on the possibility that the monetary sanction needed to deter will exceed a party's assets, and thus that he could not be deterred by threat of a monetary sanction. ... The greater this probability, the higher is the monetary sanction required to deter. ... And since the optimal sanction rises with the expected private benefits -- because parties become harder to deter -- an increase in expected harm will tend indirectly to increase the optimal sanction. ... The possible desirability of punishment of attempts according to deterrence theory does not imply that there is any advantage in punishing attempts in the way the criminal law does, namely, less severely than acts that result in harm. ... Let b be the private benefit a party would derive from committing a harmful act, where 0 <= b <= mean-b, and let h >= 0 be the harm that the act would cause; thus, a party is identified by the pair (b,h). ...