In this letter to the editor, the authors comment on an article by Busch et al (2003) on the clinical correlates of inpatient suicide, specifically the treatment implications which are said to be unclear or doubtful. The reply of Busch and her colleagues follows.