As an increasingly popular format of input, the affordances of audio-visual materials have been widely studied. Past research has provided evidence that audio-visual input combined with different captioning strategies could benefit learners in terms of vocabulary learning, listening comprehension, and the development of grammatical knowledge. However, there is a lack of research on how manipulating captioning conditions could help learners use their own linguistic resources to produce L2. Comparing the effects of three captioning techniques, L1 glossed keyword captioning, keyword captioning, and no captioning on English learners' oral and written recall of a short video, this pilot study aims to test the instruments and the data collection methods. The tentative results suggest that L1 glossed keyword captioning might have worked better in facilitating students' oral and written production of the keywords than keyword captioning and no captioning. The study also shows that L1 glossed keyword captioning might be more useful than keyword captioning and no captioning in helping students comprehend and reproduce the content of the video. Suggestions for further research on this topic are presented in the final part of this paper.