A method and a system generate a contextual visual challenge image to be presented to a user thereby to verify that the user is human. For example, an image module generates a visual challenge to be presented to a user as part of a challenge-response to verify that the user is human. A contextual background module identifies a contextual background that is contextual to a specific environment and a combiner image module combines the visual challenge and the contextual background into an image which is to be presented to the user in the specific environment, the contextual background associating the visual challenge with the specific environment.