The chance of persons successfully committing crimes they would not otherwise commit in the hope of being able to convince a jury that they could not control themselves is not plausible in reality. It is even more farfetched to think that the insanity plea weakens the resolve of some otherwise deterrable individuals who are on the brink of losing control. However, there is logical and ethical support for limiting the insanity defense to those defendants who would demonstrate a severe cognitive incapacity. Van den Haag argues that NGI defendants should be sent to mental hospitals, but given fixed sentences so that if they recovered from their mental illness they would serve the remainder of the term in a prison. Although this might remove some incentive for pleading the insanity defense, it also denies to defendants truly unable to control themselves even the limited benefits of the insanity defense -- the excuse from moral blame and presumption that they deserve treatment rather than punishment. Van den Haag feels that public safety is best guaranteed by confining NGI defendants for at least a period of time commensurate with the offense and argues as a NGI regains sanity, he or she also regains some responsibility for actions committed while insane. While the first argument has some merit, the second goes too far. It would be cruel and unusual punishment to confine an individual who has been found to be not responsible and who apparently is no longer mentally ill either in a mental hospital or a typical prison. Instead, NGI acquittees who have regained sanity after commitment and before their sentences expired should be confined, if at all, in institutional settings no more disadvantageous than our better minimal security prisons. The paper includes 19 footnotes.