Purpose – The purpose of this paper is to lay the groundwork for a narrative study of desistance that is both qualitative and quantitative. Design/methodology/approach – The review traces the strands of research that have made self-story an important theme in the study of desistance with particular reference to work since 2001. Findings – The importance of an agentic self-story in the process of desistance from crime came to prominence in the work of Shadd Maruna (1997, 2001). Since then authors have attempted to formulate: first, an integrated theoretical view of desistance incorporating agency; and second, a clinically useful understanding of how self-story is important. The clinical studies have almost always been qualitative, relying on extensive life history interviews which yield great richness of detail but few, if any, testable hypotheses. To date, such studies have not provided the empirical foundation on which to develop policy in correctional environments. Practical implications – If it is found that a measure of self-belief correlates with desistance from crime, it may be possible to devise psychological interventions to enhance and change self-belief. Originality/value – The paper proposes adding a quantitative approach to the measurement of self-concept in order to estimate the likelihood of desistance.