Developmental scientists tend to address questions about mechanism in ways that, ironically, are not especially developmental. More specifically, although we now have a great deal of data that suggest that childhood experiences have implications for human development, we know little about the time course of such effects or the dynamic mechanisms that might sustain them. Why? Because longitudinal data are rarely analyzed in a manner that can probe the developmental mechanisms by which earlier experiences are carried forward over time. In this article, we explore this paradox in detail, propose a solution, and review a set of published examples that implement the solution with a focus on the predictive significance of early maternal sensitivity. We conclude with suggestions for work in this area.