Background The anxiety disorders are robust correlates/predictors of suicidal ideation, but it is unclear whether (a) the anxiety disorders are specifically associated with suicidal ideation or (b) the association is due to co-morbidity with depression and other disorders. One means of modeling co-morbidity is through the personality traits neuroticism/negative emotionality (N/NE) and extraversion/positive emotionality (E/PE), which account for substantial shared variance among the internalizing disorders. The current study examines the association between the internalizing disorders and suicidal ideation, after controlling for co-morbidity via N/NE and E/PE.Method The sample consisted of 327 psychiatric out-patients. Multiple self-report and interview measures were collected for internalizing disorders [depression, generalized anxiety disorder (GAD), post-traumatic stress disorder (PTSD), social anxiety, panic and specific phobia] and suicidal ideation, as well as self-report measures for N/NE and E/PE. A model was hypothesized in which each disorder and suicidal ideation was regressed on N/NE, and depression and social anxiety were regressed on E/PE. Structural equation modeling (SEM) was used to examine the unique association of suicidality with each disorder, beyond shared variance with N/NE and E/PE.Results The hypothesized model was an acceptable fit to the data. Although zero-order analyses indicated that suicidal ideation was moderately to strongly correlated with all of the disorders, only depression and PTSD remained significantly associated with suicidal ideation in the SEM analyses.Conclusions In a latent variable model that accounts for measurement error and a broad source of co-morbidity, only depression and PTSD were uniquely associated with suicidal ideation; panic, GAD, social anxiety and specific phobia were not.