Background and methods Suicidal events during antidepressant treatment are a major factor affecting antidepressant prescribing. In this perspective, regulatory bodies in the US as well as in Europe issued a “black box warning” regarding the risk of suicidal events related to the use of antidepressants, not only in children and adolescents but also in adults. Although the clinical risk factors for such events are well known, unfortunately they do not help to predict who will have a suicidal event during antidepressant treatment and who will not. Recently, pharmacogenomic studies have attempted to use genetic variants to predict individual susceptibility to treatment-related suicidal ideation. This thesis reviews the literature about genetic predictors of antidepressantrelated suicidal events. In this perspective, I will describe more in details two studies I had the chance to be involved in: the Genome-Based Therapeutic Drugs for Depression (GENDEP) and the Geneva Outpatient Depression Study (GODS). Results: Several genetic predictors have been highlighted, the majority of which is relate to common mechanisms of antidepressant action, some of them overlaping with genetic predictors of antidepressant response: genes involved in the neurotrophic and synaptic plasticity systems(CREB1, BDNF and its receptor NTRK2), in noradrenergic and serotoninergic systems,(ADRA2A and 5-HTR1B), in glutamatergic system (GRIA3, GRIK2 and GDA), in the inflammatory and HPA-axis systems (IL28RA and FKBP5), in the control of the blood–brain barrier (ABCB1) and in other brain functions (PAPLN, APOO, KCNIP4 and ELP3). Discussion: These results underscore the importance of multiple genetic factors in the prediction of antidepressant induced suicidal events, which certainly underscores the multifactorial nature of this trait as well as its overlap with other phenotype such as treatment response. Although some of these genes may be of interest in predicting antidepressant-induced suicidal ideation they still need to be validated in better phenotypically designed samples. Several methodological factors are indeed responsible for the problems involved in implicating these findings in the causation of a clinically relevant phenotype. These include: discrepancies between studies in defining phenotypes with several different thresholds used to establish significant suicidal ideation, the use of scales not truly designed to measure suicidal ideation and the paucity of true suicidal events (suicide attempts and/or completion) in 6 pharmacogenomic studies. In conclusion, pharmacogenomic studies are far from fulfilling their promise. There is a need for future pharmacogenetic studies targeting events that are clinically significant in order to find associated variants that will help clinicians to improve their treatment strategies. While awaiting these genetic predictors, clinicians need to bear in mind that all studies in this field including GENDEP and GODS support a beneficial effect of antidepressants on suicidal ideation. This should therefore encourage them to prescribe antidepressant medication even in patients with suicidal ideation.