Early interventions for psychosis have recently received a great deal of interest. This is because research findings have indicated the importance of duration of untreated psychosis (DUP) in influencing future prognosis. Most interventions have so far centered around the early identification of first episode cases in an attempt to minimize the DUP. A further development of the early intervention strategy aims to identify high-risk cases prior to the onset of psychosis. Treatment of this at risk group could potentially prevent the onset of psychosis. In this paper we describe three cases from a randomized controlled trial designed to test the feasibility of a cognitive therapy intervention, to prevent the onset of psychosis. These are presented with formulations based on a recent cognitive model conceptualizing the onset of psychosis. Implications of this approach for future research and clinical practice are also discussed.