Students learn large amounts of information, but not all of it is remembered after courses end -- meaning that valuable class time is often spent reviewing background material. Crucially, laboratory research suggests different strategies will be effective when reactivating previously learned information (i.e. "marginal knowledge"), as opposed to learning new information. In two experiments, we evaluated whether these laboratory results translated to the classroom. Topics from prior courses were tested to document which information students could no longer retrieve. Half were assigned to a not-tested control and half to the intervention; for these topics, students answered multiple-choice questions (without feedback) that gave them the chance to recognize the information they had failed to retrieve. Weeks later, students completed a final assessment on all topics. Crucially, multiple-choice testing increased the retrieval of previously forgotten information, providing the first classroom demonstration of the reactivation of marginal knowledge.