Environmental challenges would seem to require scientifically informed predictions in order to effectively resolve. However, sustainability scientists have exhibited a tendency to reject the concept of “prediction” as a useful goal. In this paper we explore how many sustainability scientists have come to discard prediction as a meaningful concept, and how they conceptualize what they are doing instead to guide decision formation or policy-making goals. We will suggest however that rejecting prediction—canonically construed—by an applied or management-orientated field such as sustainability science has both ethical and epistemic costs, and there are benefits to striving for prediction even if this is difficult to achieve.