摘要:Tests have been vastly used for the assessment of learning in educational contexts. Recently, however, a growing body of research have shown that the practice of remembering previously studied information (i.e., retrieval practice) is more advantageous for long-term retention than restudying that same information; a phenomenon often termed “testing effect”. The question remains, however, whether such practice can be useful to improve learning in actual educational contexts, and whether in these contexts specific types of tests are particularly beneficial. We address these issues by reviewing studies that investigated the use of retrieval practice as a learning strategy in actual educational contexts. The reviewed studies adopted from free-recall to multiple-choice tests, and involved from elementary school children to medical school students. In general, their results are favorable to the use of retrieval practice in classroom settings, regardless of whether feedback is provided or not. Importantly, however, the majority of the reviewed studies compared retrieval practice to repeated study or to “no-activity”. The results of the studies comparing retrieval practice to alternative control conditions were less conclusive, and a subset of them found no advantage for tests. These findings raise the question whether retrieval practice is more beneficial than alternative learning strategies, especially learning strategies and activities already adopted in classroom settings (e.g., concept mapping). Thus, even though retrieval practice emerges as a promising strategy to improve learning in classroom environments, there is not enough evidence available at this moment to determine whether it is as beneficial as alternative learning activities frequently adopted in classroom settings.