Stealth assessment is intended to not only measure important competencies, but also to support their development during gameplay or within other types of immersive learning environments. It uses evidence-centered assessment design to create the models (Mislevy, Steinberg, and Almond 2003), in conjunction with a statistical scoring and accumulation approach and educational data mining techniques to capture and analyze real-time data and dynamically measure students' learning. This chapter describes the background and benefits of stealth assessment, reviews a growing body of empirical studies that support stealth assessment in different digital games, and describes how stealth assessment can be embedded in a game via a ten-step iterative process. Finally, the chapter discusses some of the hurdles and obstacles to overcome and current efforts on developing various types of learning supports to accompany stealth assessment in a game. [This chapter was published in: J. M. Spector (Ed.), "The Routledge Encyclopedia of Education" (p1-9). London, UK: Taylor & Francis.]