Disclosed are a method, an apparatus, and a system for generating a posture virtual image, and a virtual reality device, to resolve a technical problem that the virtual image generated based on the prior art is not realistic enough. A method for generating a posture virtual image of a user includes: separately obtaining initial images of key regions of a first part of a user; extracting position information of feature points from the initial images; integrating the initial images according to position information to generate a first part image; and generating a virtual image according to the first part image. The initial images of the plurality of key regions can represent the first part of the user accurately, thereby a real feature of the first part of the user is accurately presented, and a real-time experience of the user during interaction is improved.