A keystroke pattern belongs to dynamic (behavioral) biometric characteristics describing the subconscious actions familiar to a user. It characterizes the passphrase entering dynamics. The keystroke pattern recognition is based on a comparison of the pattern in question with a list of pattern templates stored in a memory of a computer. Thus, the problem of user identification reduces to a problem of pattern recognition. Keyboard usage analysis does not require special hardware and personnel for installation and further support. It allows the process of covert authentication and is transparent to end user, i.e. does not cause any inconvenience. Classic statistical approach to the keystroke pattern recognition (while typing several key words) reveals a number of interesting features: the dependence of character combinations in a typed word, a deep connection between a set of individual characters, a “delay” presence when typing characters.