The current study examines associations between community violence exposure (CVE) and internalizing and externalizing symptoms. The emotional desensitization hypothesis states that moderate levels of CVE will be associated with increased internalizing symptoms, whereas high levels of CVE will be associated with decreases in internalizing distress. The current study extends prior research by examining whether patterns of emotional desensitization differ based on the route of exposure – either being personally victimized or witnessing community violence. Further, associations examined the influence of interpersonal proximity to the victims of violence – that is, whether individuals are victims themselves (most proximal), witness violence against known individuals, or witness violence against strangers (least proximal). Participants include 714 undergraduate students (Mage = 20.5 years, SD = 3.9) who completed an online survey of CVE, internalizing and externalizing outcomes, and cognitions supporting aggression. Results revealed significant moderation by interpersonal proximity. Specifically, curvilinear associations between CVE and internalizing symptoms, reflecting emotional desensitization, were stronger at lower levels of proximity, whereas positive linear associations emerged at higher levels of proximity. Associations between CVE and more externalizing behaviors were stronger at low, compared to high, levels of proximity. Internalizing symptoms and cognitions supporting aggression partially mediated the association between CVE and externalizing behaviors. Further, latent profile analysis revealed that distinct groups of individuals were classified based on cumulative frequency of CVE and types of violent experiences. Thus, route of exposure, relationship to the victims of violence, and type of violence are each salient characteristics of CVE that may relate differentially to adjustment.