Clinical prediction models of outcome for patients with a psychotic disorder do not exist at present. However, since there is abundant evidence of factors influencing outcome, it should be possible to realize our main goal: To develop a prediction model for recovery outcome for the individual patient that also indicates how changing certain factors can influence it in a favorable way. What is meant by ‘favorable’, is part of the project: Patients and psychiatrists together will define outcome variables that are important to them. We will apply advanced statistical techniques such as machine learning to the large (N=2000) longitudinal multisite datasets we have at our disposal and validate the resulting prediction tool in new patients. Insight in how changing (lifestyle) factors may improve a patient’s personalized outcome will increase the patient’s opportunities for self-management, enable tailored interventions, reduce societal costs and improve the lives of those affected by this severe illness.