Conducted with a community sample, this study first tested the hypothesis that the constant association of Non-Suicidal Self Injury (NSSI) with various psychopathological states may be overrepresented because many of the studies reporting these associations used clinical samples. Secondly, the study aimed to test the emotional regulation model of NSSI by exploring the functions, the affective antecedents and consequences of NSSI episodes and to understand this emotion regulation role of NSSI in the light of Attachment Theory by differentiating self-injurers and non-self-injurers on relevant romantic attachment dimensions. A third purpose of this study was to test the hypothesis that individuals currently performing NSSI could be differentiated on these dimensions from those who had ceased engaging in NSSI. Pursuing these purposes, 518 university students (171 males and 347 females), aged 17 to 62 years old completed the Self-Injury Questionnaire – Treatment Related (Claes & Vandereycken, 2007), the Brief Symptom Inventory (Derogatis, 1982) and the Adult Attachment Scale (Collins & Read, 1990). Individuals with NSSI scored significantly higher on all BSI subscales (all p < .001). Results also revealed the existence of significant differences between participants with and without NSSI on Anxiety (Z = –2.92, p < .01) and Comfort with Proximity (Z = –3.18, p < .01), and significant differences between past self-injurers and current self-injurers on Trust in Others (Z = –2.40, p < .05). These results are discussed by linking NSSI and Attachment Theory literatures.