The common idea of video analysis in physics education so far is that students record the motion of an object using a video camera, transfer the video on a computer and analyze it with a suitable software. In our approach students perform video analysis time-efficiently using only tablet computers. We present the results of a study which examines learning effects of tablet PC-supported video analysis in high school physics courses in a pre-post-test design with treatment and control groups for two essential topics of mechanics, the uniform motion (N = 109 matched samples) and the accelerated motion (N = 70 matched samples). The results show that the treatment leads to significantly higher learning gain regarding conceptual knowledge in comparison to traditional teaching without video-based motion analysis, especially for the cognitively more demanding topic. We discuss the results in the context of learning theories and reveal implications for future research.