Predictive capacity is needed to anticipate the consequences of global change. Along with the computational challenges inherent in accounting for uncertainly in models of ecological and physical processes related to global change, we face educational challenges related to developing the intellectual capital for thinking with uncertainty in mind. Hand in hand with promoting the kinds of research needed to advance the evolving science of ecological forecasting, we need to set an education agenda for developing and enhancing computational literacy of current and future ecologists, managers, and policymakers. Key elements of an educational agenda are (1) defining the audiences and what each needs to know and be able to do; (2) developing curricula and pedagogical strategies in which thinking skills and conceptual understanding are, by design, linked; (3) addressing training needs of faculty to teach effectively about uncertainty to these different audiences; and (4) creating and implementing assessment tools to explore the impact of programs designed to train ecologists to think with uncertainty in mind. Assessing the extent to which new training models have impacted forecasting research and the conceptual understanding of ecologists is an important topic for future scholarship in education, and a natural area for collaboration between forecasting and education researchers.