Diagnoses have heterogeneous outcomes, varying from good to extremely poor. There is a need to single out an ultra-high-risk group of individuals who have illnesses that might well end unfavourably or who might later develop serious psychopathology.<br/> AIM: To devise a screening instrument that can identify a group of individuals who run a very high risk of developing a first-episode psychosis, and to create a type of intervention that can modify the course of the illness.<br/> METHOD: We developed a short screening instrument (PQ-16) and were able to ascertain its predictive value. We also tested an intervention that could influence risk factors and deal with emerging symptoms thereby achieving a better outcome for the patient.<br/> RESULTS: We developed a two-step detection instrument with a positive predictive value of 44%. The intervention, involving cognitive behavioural therapy for ultra-high-risk patients, was effective and led to a risk reduction of about 50%. Using the ultra-high-risk group of patients, we were able to model three prognostic profiles, each carrying a 4%, 13%, and 70% risk of subsequently developing psychosis. The intervention was cost-effective, reducing the financial burden on the health care services and on society as a whole.<br/> CONCLUSION: Prognostic modelling and proactive intervention can achieve improvements in health at lower costs.