BackgroundIdentifying patients at risk for a suicide attempt (SA) is critical in adolescents with mental disorders. The current study aimed to 1) examine whether personality dysfunction (PD) is associated with previous SA, 2) explore the incremental utility of PD over psychiatric disorders in modeling previous SA. MethodsThe sample comprised of n = 498 adolescent patients (mean age = 15.41 years, 79.12 % females, inpatient 48.8 %, outpatient 51.2 %). SA in the past year, PD according to the alternative DSM-5 model for personality disorders, and psychiatric diagnoses were assessed using semi-structured interviews. Logistic regression and principal component analysis examining the associations and specific patterns of PD and SA in the past year were conducted. Hierarchical (stepwise) logistic regression was applied to investigate the incremental utility of PD over that of psychiatric diagnoses to identify individuals with SA in the past year. ResultsIncluding all facets of PD revealed a significant model with SA in the past year as outcome (χ2(12) = 106.65, McFaddens Pseudo-R2 = 0.17, p < 0.01). Adding PD to the model explained a significant amount of variance in past SA over that of psychiatric diagnoses (Pseudo-R2 = 0.18, Wald χ2 = 43.05, p < 0.01). LimitationsAs we only studied past SA and due to the cross-sectional design, no conclusion regarding the prediction of future SA can be drawn. DiscussionPD should routinely be assessed in adolescent patients since individuals with PD are more likely to have attempted suicide even when controlling for comorbid psychiatric disorders. PD may represent an important target for intervention in those with suicidal thoughts and behaviors.