Persons with mood disorders are at the greatest risk for suicide, with 50–70% of persons who kill themselves doing so during an episode of depressive illness. Melancholic patients who are agitated and anxious, psychotic, or who have been hypomanic are at the greatest risk. Studies of suicide, however, do not typically identify the melancholic patients in their samples, and so the evidence for risk is mostly indirect. Patients who are severely depressed, or who have abnormal hypothalamic–pituitary–adrenal functioning are likely to be melancholic, but studies associating these factors of increased risk rarely define the form of depressive illness. An exception is a study that compared suicide attempts in a large sample of melancholic and non-melancholic patients. After controlling for severity and baseline characteristics as covariates, the melancholic patients were more likely to have had prior serious suicide attempts and to make more attempts during the follow-up period. Although the cited studies refer to severe depression, it is likely that the majority of patients meet criteria for melancholia.

In early estimates of suicide risk, 15–19% of depressed patients committed suicide.