Various studies have shown that video-based learning by explaining to a fictitious audience can be an effective learning strategy for promoting multiple knowledge domains such as memory, comprehension and knowledge transfer. However, field studies testing the effectiveness of this learning strategy in an applied setting are rare. The present study examines the effectiveness of the one-take video (OTV) approach on undergraduate students' learning performance. The OTV method involves users recording short oral presentations without any editing and with the support of handwritten visualisations (video-based learning by explaining). To test the learning outcomes, 218 undergraduate teaching students for special educational needs were randomly assigned to two test groups (OTV and explaining in writing). After that, they completed three study tasks throughout the semester, each followed by immediate and delayed knowledge tests. The results for the OTV group show that students achieved significantly better results in the immediate memory test the more handwritten visualisations they used, but not the more often they repeated the video recordings. Analyses of variance revealed that the OTV group outperformed the writing group in terms of memory performance in the immediate test but not in the delayed posttest. The OTV group also significantly outperformed the writing group in both posttests in the transfer domain. No significant differences were found in the comprehension test scores.