BACKGROUND: Suicidal risk is often unrecognized in emergency department (ED). We aimed to assess its prevalence in patients with long-term conditions (LTCs) attending an ED and to test whether gender differences influence suicidal risk assessment, using the diagnostic accuracy properties of the Risk Assessment Suicidality Scale (RASS).METHODS: The RASS was administered to 349 patients with diabetes, COPD and rheumatic diseases visiting an ED. The MINI interview was used as the criterion standard. ROC curve analysis was performed to determine the optimal RASS cutpoint for suicidal risk separately for males and females. Somatic (PHQ-15) and depressive (PHQ-9) symptoms were also assessed and factors associated with suicidal risk across gender were determined in hierarchical regression models.RESULTS: The prevalence of suicidal risk according to the MINI was 22.9%; 16.6% of patients were at low, 5.1% at moderate, and 0.9% at high risk. At an optimal cutpoint of 270, RASS had 81.3% sensitivity and 81.8% specificity. The optimal RASS cutpoint for females (340) was double the cutpoint for males (175). Somatic symptom burden was associated with suicidal risk in both sexes but it became non-significant after depressive symptoms were taken into account; suicidal risk was also associated with history of depression in females and lower income in males.CONCLUSION: There is a high prevalence of suicidal risk in patients with LTCs attending the ED. As the optimal RASS cutpoint for females was double the cutpoint for males, clinicians should bear in mind gender differences when assessing for suicidal risk in the ED.