There has been considerable discussion in the literature on the differences between criteria for involuntary commitment that are based on dangerousness and criteria based on need for treatment. A number of states have adopted clinical criteria, and other state legislatures are actively considering them. Some libertarians argue that dangerousness is constitutionally required if a person is to undergo the loss of liberty involved in commitment. Citing widely publicized data from the state of Washington, they predict that a return to clinical criteria would result in a deluge of inappropriate commitments. Some clinicians counter that use of clinical criteria would result in selection of a much more appropriate clinical population and point to research indicating that strict observation of the need-for-treatment provisions of the APA model commitment statute would actually decrease the number of commitments. The author examines state hospital admission and census data from eight states that added need-for-treatment criteria to their commitment codes between 1975 and 1990 and argues that the data indicate that there is little reason to believe that such changes would result in the deluge of admissions predicted by the critics.Abstract Teaser