Recent reviews of the desistance literature have advocated studying desistance as a process, yet current empirical methods continue to measure desistance as a discrete state. In this paper, we propose a framework for empirical research that recognizes desistance as a developmental process. This approach focuses on changes in the offending rate rather than on offending itself. We describe a statistical model to implement this approach and provide an empirical example. We conclude with several suggestions for future research endeavors that arise from our conceptualization of desistance.