The diagnosis of Complicated Grief (CG) is being proposed for inclusion in DSM–5. As such, it has been the focus of several studies purporting to build evidence of its validity and its conceptual and statistical distinction from Posttraumatic Stress Disorder (PTSD) and depression (Bonanno et al., 2007; Golden & Dalgleish, 2010; Prigerson et al., 1995b). However, previous research has focused predominantly on bereavement caused by non-violent means (e.g., prolonged terminal illness). This study attempts to explore the nature of CG among a sample of students who survived a mass shooting. Exploratory and confirmatory factor analytic procedures were used to examine and confirm the factor structure of CG as reported in previous studies (e.g., Boelen & van den Bout, 2005; Dillen, Fontaine, & Verhofstadt-Denève, 2008). A refined CG scale was then used as a criterion to demonstrate how different types of traumatic exposure contribute to symptoms of CG and/or posttraumatic stress (PTS). It was hypothesized that exposure items related to bereavement would be more related to CG than to PTS, whereas direct exposure to the shootings would be more closely related to PTS than to CG. Results of exploratory and confirmatory factor analyses supported CG as a unitary construct distinct from PTS and from anxiety/depression. Logistic regression results demonstrated that bereavement status was significantly predicted by CG but not PTS: Participants who scored 1 standard deviation above the mean on the CG scale were 14.64 times more likely to have been bereaved than were those who scored at the mean. SEM analyses were used to provide an additional test of this hypothesis. The final model had acceptable fit as assessed by RMSEA = .046, CI = .043–.049, SRMR = .048, and CFI = .990; however, the Satorra-Bentler Scaled Ï 2 = 1507.82, df = 589, p < .001, did not support the model. Overall, results of SEM suggested that interpersonal loss (i.e., whether a friend was killed, injured, or escaped from the shootings) predicted CG but not PTS, whereas perceived threat predicted both CG and PTS.