This chapter focuses on the theoretical foundations of virtual reality (VR) for language learning and their empirical practices. Two main theory umbrellas, sociocultural theory and embodied cognition, form the trunk of the theoretical foundations of this chapter. The former emphasizes the importance of social interaction in the development of cognition, while the latter argues that the representation of knowledge is grounded in a person’s experiences of interacting with and perceiving the environment, which involves bodily sensation, perceptions, and actions. Although there are plenty of theories supporting foreign language learning, the abovementioned theories are chosen because they give prominence to experience-oriented learning and support kinesthetic learning. Experience-oriented learning refers to situated cognition and benefits learners’ pragmatic linguistic skills. By involving foreign language learners in VR, the three features of VR (immersive, interactive, and imaginary) match the essential components of successful language learning, i.e., learners’ active involvement in an authentically meaningful and social interaction. Therefore, the two theories mentioned above, sociocultural theory and embodied cognition, well bridge the features of virtual reality and the essential components of successful language learning. Following the description of the theoretical foundations, the essential components of implementing the abovementioned theories in empirical practices are elaborated. Based on the theoretical foundations and the implementation principles, five categories of empirical practices are introduced, including social connection, game-based learning, self-exploration, cooperative task-based learning, and learning by creation. Finally, some reminders and suggestions are given to conclude the chapter.