Data were obtained from a multistate, longitudinal study of insanity defense reforms. Four States (California, Georgia, New York, and Ohio) were selected because they had enacted insanity defense reforms. The remaining three States (New Jersey, Washington, and Wisconsin) were selected as control States since they had not enacted insanity defense reforms during the study period. Results indicated that offense seriousness was a more important factor than mental disorder in determining length of confinement for persons found not guilty by reason of insanity. Persons found guilty were more likely to be released without ever having been confined than persons acquitted by reason of insanity. Persons acquitted by reason of insanity were typically confined to mental hospitals until such time as it could be determined they were no longer dangerous. Implications for invoking offense seriousness as a primary criterion in assessing dangerousness are discussed. 20 references and 5 tables