This teacher-researcher study explored the manner in which students created video compositions in a secondary English language arts media studies program. A review of research literature indicates fundamental differences between print and video compositions, which include modality of representation, task setting, and curricular role. Another difference is in the description of models of composition in studies of both print and video. Although print research details stage and recursive aspects, studies of video used models and/or terminology that were expressed in linear and stage terms. This yearlong, classroom-based study examined the video composition processes of three case study groups (n=3, n=3, n=4) within the context of both their classroom (n=19) and the entire Communications program (n=82). This study used methodology from studies of print composition (think aloud and retrospective think aloud protocols) in order to more fully explore the processes students used to compose their videos. The study found that video composition is a complex, recursive process that allows for sequential multimodal representation of thoughts and ideas. Four areas are addressed: video allows for the expansion of compositional choices, demonstrates the verisimilitude of students' initial concept to videotaped image, highlights the visuality in students' re-presentations of ideas, and provides research methodological considerations. A model was developed to describe the recursive stage and phase aspects of the students' video compositions. (Contains 5 tables and 3 figures.)