ABSTRACT Offenders with Severe Mental Illness: The Role of Reactive and Proactive Violence in Recidivism Peter Simonsson, MSW, LCSW Phyllis Solomon, Ph.D. Statement of Problem. Probationers and parolees with severe psychiatric disorders have higher rates of violent recidivism than offenders without mental illness. Extant research suggests that the most effective way to examine violence is to delineate violence into two types: reactive and proactive. The purpose of this mixed-methods dissertation was to examine the relationship between severe psychiatric disorders and violence type and whether static, dynamic, and situational factors explained type of violence among probationers/parolees with/without psychiatric disorders. Methods. Mixed-methods using a longitudinal design for the quantitative portion and semi-structured interviews for qualitative portion. Probationers/parolees were recruited from the probation department in Philadelphia, PA. Ninety-eight probationers/parolees meeting eligibility criteria were included in the study. Twenty-four offenders of the total sample also participated in semi-structured interviews. The study examined: 1) the association between reactive and proactive violent recidivism and probationers/parolees with and without severe psychiatric disorders;; 2) the extent to which the presence of severe psychiatric disorders predicts violent recidivism; 3) the extent to which dynamic (substance use, social isolation, social supports), and static (juvenile adjudication, violent priors, criminal history, childhood trauma) factors predict proactive and reactive violent recidivism among probationers/parolees with and without psychiatric disorders; and 4), how situational factors (relationship dynamics, employment, housing) influence reactive and proactive violence. Quantitative data were analyzed using Multivariate Logistic Regression and Chi-Square and qualitative data with a thematic approach. Results. Hypotheses were partially supported, and certain trends could be detected. The presence of a severe psychiatric disorder did not increase the risk for violent recidivism; however, when analyzing type of violent recidivism, severe psychiatric disorders were associated with reactive violent recidivism. Dynamic factors predicted reactive, but not proactive violent recidivism. Static factors predicted proactive violent recidivism. The qualitative results highlighted how reactive violence emerges out of family conflicts, whereas proactive violence tends to occur in the context of the underground economy. Implications. Probationers/parolees with/without severe psychiatric disorders engage in different violence types. Since probationers/parolees with severe psychiatric disorders primarily engage in reactive violence targeting family members, these findings offer directions for the design of interventions to prevent violence.