AbstractBackground:Suicide has been recognised as one of the major causes of premature death in psychosis. However, predicting suicidal behaviour (SB) is still challenging in the clinical setting and the association of neurocognition with SB in psychosis remains poorly understood. This study aimed to investigate the role of neurocognitive performance as predictor of SB. Also, we sought to explore differences in the evolution of clinical and neurocognitive functioning between participants with/without history of suicide attempts (SA) over follow-up period.Methods:The sample of the study is composed by 517 patients. Sociodemographic, clinical, functional and neurocognitive measures were evaluated at baseline as well as 1-year and 3 years after first episode of psychosis. Bivariate and multivariate analyses explored the influence of these variables as putative baseline predictors of SB. Repeated measures analyses of variance tested differences in clinical and neurocognitive outcomes at 1- and 3-year follow-up.Results:Global cognitive functioning (GCF) (OR = 1.83, 95% CI = 1.25–2.67) and severe depressive symptoms (OR = 1.17, 95% CI = 1.07–1.28) predicted SB. Longitudinal analyses revealed that patients with SB at follow-up presented with higher levels of remission in terms of positive psychotic symptoms and depression. In addition, those with a history of SB had worse GCF and visual memory than those without such antecedents.Conclusions:GCF was found to be the most robust predictor of SB along with severe depressive symptomatology. Hence, poorer cognitive performance in FEP appears to emerge as a risk factor for suicidal behaviour from early stages of the illness and a comprehensive neurocognitive assessment may contribute to risk assessment.