Background: Instructional videos are increasingly employed in a physical education domain in order to enhance motor learning and self-motivational beliefs. However, evidence showed that videos are generally ineffective because they often create extraneous loads due to the information transience and passive processing of information. Purpose: The current study investigates the effects of controlling the display of video on learning outcomes -- expressed in terms of recall and transfer scores -- and situational interest in physical education settings. Method: 60 tenth-grade students (M[subscript age] = 15.90 years, SD[subscript age] = 1.32 months; 30 boys and 30 girls) were instructed to study using either a continuous video (presented without pauses), a system-controlled video using predefined segments or a self-controlled video where the learner uses 'stop' and 'play' keys. Then, they were asked to perform the learning tests (i.e. game comprehension test and game performance test) and evaluate the situational interest of the learning materials (i.e. attention demand, challenge, exploration intention, instant enjoyment, novelty). Findings: The results from one-way ANOVAs revealed that students performed significantly better on both recall and transfer post-tests when the video was either system-controlled or self-controlled, which suggested that studying a controlled video could decrease irrelevant cognitive load and so yield better learning performances. Concerning the situational interest variable, students in the self-controlled condition scored significantly higher on almost all dimensions of situational interest. Conclusion: This study indicates that self-controlled video, as applied in this study, can be integrated in a physical education context to foster learning and enhance situational interest. The discussion advances several options for enhancing the effectiveness of instructional videos during physical education lessons.