Retrieval-based learning, using tests for content review, frequently proves more effective for knowledge retention compared to alternative methods. Extensive research has explored this with older students, often in contrast to more passive techniques like rereading or note rewriting, typically focusing on vocabulary content, in non-classroom settings and assessing knowledge retention within a week. In this study we conducted a classroom experiment to compare the efficacy of retrieval-based learning to another active method of reviewing content, namely discussion, in promoting long-term knowledge retention among primary school science students. Additionally, we assessed the students' perceptions of these reviewing methods. A total of eighty-one primary school students participated in the study, which encompassed a lesson conducted at a zoo, followed by a review of the lesson content using either retrieval-based learning or discussion. Our findings indicated no impact of the method of reviewing on knowledge retention after either one week or four weeks. Nevertheless, notable effects were observed in terms of motivation, as students who employed retrieval-based learning for reviewing reported heightened levels of interest and enjoyment, along with perceiving greater benefits from the process. Additionally, boys in the discussion group reported increased curiosity, in contrast to girls who reported an inverse pattern.