An apparatus for context awareness includes: a voice-based recognition unit that recognizes a user's emotional state on the basis of a voice signal; a motion-based recognition unit that recognizes the user's emotional state on the basis of a motion; a position recognition unit that recognizes a location where the user is positioned; and a mergence-recognition unit that recognizes a user's context by analyzing the recognition results of the voice-based recognition unit, the motion-based recognition unit, and the position recognition unit. Accordingly, it is possible to rapidly and accurately accidents or dangerous contexts caused to a user.