Many studies have demonstrated that infants can attribute goals to observed actions, whether they are presented live by familiar agents or on a computer screen by abstract figures. However, because most, if not all, of these studies rely on the repeated action presentations typical of infant studies, it is not clear whether infants are simply recognizing the completed action as goal directed, or whether they can productively infer a not-yet-achieved outcome from an ongoing action. We investigated this question by presenting 13-month-old infants with a single animated chasing event. Infants looked longer at the outcome of this action when, given the opportunity, the chaser did not catch the chasee than when it did. Crucially, this result was dependent on whether the action could be construed as efficient with regard to this goal state. This finding suggests the ability to infer the goal of an ongoing novel action and illustrates the productivity of 1-year-olds' action understanding.