Personalised or precision medicine aims at tailoring medical decisions and interventions to the individual patient based on their predicted illness course or risk of disease. This would involve, for example, matching individual patients with the most effective psychological or pharmacological treatments while minimizing the risk of adverse events. This could have many benefits for patients and society and reduce the healthcare costs arising from the use of inappropriate or suboptimal treatments. In keeping with this, and given the global burden and the clinical heterogeneity of psychotic disorders, early identification of psychosis patients at higher risk of adverse outcome remains a priority. This is particularly relevant in those with early-onset psychosis (EOP; i.e onset before age of 18), as their neurobiological and psychosocial development is not yet complete. In this symposium we aim to (i) perform a comprehensive review of the literature to date on the prediction of poor outcomes in EOP and (ii) present novel research in this field based on longitudinal prospective studies targeting the search for such predictors.