Biometrics–based authentication utilizes something that a user is (physiological biometrics) or something that a user does (behavioral biometrics) within the identity verification process. Typing biometrics belongs to the behavioral biometrics category. Authentication systems based on typing biometrics verify the user’s identity using the rhythm and speed of his/her typing. In this thesis, based on examples from related work, the complete authentication process based on typing biometrics is described in detail, and the existing TypingDna service is additionally tested experimentally. A system has been developed to collect individual typing patterns (in both the desktop and the mobile domain), as well as to forward them to the TypingDna API. An experimental study of the efficiency of authentication based on typing biometrics was conducted, with special emphasis on the typing pattern type, the length of the verification text, the target domain, and the way the device is held within the mobile domain. The analysis of the obtained results revealed that the authentication with the TypingDna service in the desktop domain is superior to that in the mobile domain. Mobile domain support, that needs to involve different ways of holding the device when working with emulated keyboards, requires further improvements to become a viable option. This is in line with the current version of the corresponding service, which is still in beta phase.