As part of the research protocol, each subject was administered the Clinical Analysis Questionnaire (CAQ), a factor-analytically derived psychometric personality assessment method. The 16-PF of the CAQ allows the subject personality to be quantified in terms of 16 "normal personality" constructs or domains. The findings reported in this article concern only the 16-PF portion of the test. Pearson Correlation coefficients were calculated for the 16-PF factor scores in each drug abuse category. Few subjects used only one drug, so the data for each subject were entered into the correlation analysis for the category of each drug involved at the time of his crime. As a consequence of the categorical approach to analyzing this data set, the sample sizes within each drug category were necessarily small; for this reason, these results should be viewed as preliminary findings. The study found that psychostimulant abusers (hallucinogens, cocaine, and amphetamines) were distinguishable from other violent criminals in retaining paranoid and schizotypal patterns of responding in their 16-PF testing. This supports the hypothesis that this drug abuse typology results in paranoid personality trait structure. These preliminary findings may be relevant to public health strategy and to considerations of criminal culpability, policing, and corrections. 3 tables, 1 figure, and 23 references