Suicidal ideation (SI) is a major suicide risk factor; therefore, it is crucial to identify individuals with SI. Discrepancies between the clinicians and patients’ estimation of SI may lead to under-evaluating the suicide risk. Yet, studies on discrepancies between self- and clinician-rated SI are lacking, although identifying the patients’ sociodemographic and clinical characteristics associated with such discrepancies might help to reduce the under-evaluation risk. The aims of these studies were to determine: 1) Association between discrepant self- and clinician-rated active SI with clinical characteristics and suicide attempt (SA) risk; 2) which SI assessment (self or clinician) predicted risk of SA. The first study was performed on two large, prospective, naturalistic cohorts (GENESE and LUEUR) of French adult outpatients with unipolar depression treated and followed for 6 weeks. Active SI presence was assessed and defined with a score to the suicidal item of the Montgomery–Åsberg Depression Rating Scale (MADRS-SI) ?3 either self- and/or clinician-rated. Discordant SI was defined as SI detection by only one of the two evaluators (patient or clinician) and concordant SI as SI detection in both self- and clinician-rated questionnaires. The second study was performed on patients recruited by the French network of FondaMental expert centers for bipolar disorders, FACE-BD cohort. Presence of active SI was defined by a score ?2 for item 12 of the Quick Inventory of Depressive Symptomatology-Self Report and/or ?3 for MADRS-SI. In the three cohorts, 49.3% (GENESE), 34% (LUEUR) and 74.4% (FACE-BD) patients had discordant SI. Clinical characteristics, especially current depression levels, were more severe in patients with concordant SI than in patients with discordant SI. Risk of SA during follow-up was higher in patients with current SI (concordant and discordant) than in patients without SI and in the concordant than in the discordant group. Finally, prediction of the risk of SA was comparable by the two ratings (clinicians and patients). Patients with SI (concordant and discordant) have more severe clinical characteristics. However, in both studies patients with SI only self-rated have less severe levels of depression than patients with concordant SI, while risk of SA during follow-up was close to those with concordant SI. It seems that clinician may under-evaluate SI presence in patients with low depression levels.