This study examined one part of the criminalization thesis, which holds that the dangerous-ness standard of reform civil commitment law has led to the frequent arrest of mentally ill persons. It followed a large statewide sample of civil commitment candidates for 6 months through arrest records to observe their number and type of arrests. It found that: ex-candidates were seldom arrested; over half of arrests were accounted for by a few with multiple arrests; the nonviolent and those released were more likely to be arrested than the violent and those committed; and charges were more often for the less serious offenses, but not predominantly for nuisance offenses