We introduce TrajecStories, a novel trajectory-based visualization technique for authoring spatial summaries from first-person motion videos. Such videos are increasingly available for outdoor activities (e.g., biking to work) and capture rich information about one’s surroundings (e.g., path, obstacles) and conditions (e.g., weather). However, effectively communicating the wealth of information embedded in these videos, as well as trajectory data (e.g., speed), biometric (e.g., heart rate), and subjective information (e.g., exhaustion), remains a challenge at the intersection of video processing and data visualization. TrajecStories enables the communication of such data with textured trajectories as the main visual encoding, along with incremental data binding to include additional data based on their importance in a coherent story-oriented way. We implemented an interactive design probe to explore the underlying design space and showcase how TrajecStories can be leveraged to yield spatial summaries for various application domains. We also report on early user feedback on the expressivity of TrajecStories with personal data collections. Finally, we discuss the research agenda and the technical challenges to better include this technique in the current visualization pipeline and design processes.