According to this method of temporal segmentation of a video sequence (200) containing one or more markers, the analysis of the content of these markers makes it possible to segment the video sequence into a plurality of video sub-sequences.