Recently, IT specialties have become one of the most demanded specialties in the world labor market. Simultaneously the traditional teaching in conditions of mass production, even with a professional teacher, has a significant drawback – the fundamental impossibility of adapting to each student. Since the 60s of the twentieth century, researchers worldwide have been developing various computer-tutoring tools that have, less or more, adaptive functions. Nevertheless, the task of the perfect computer tutor development is still far from being solved. The article's research subject is the process of student requests analyzing during intelligent computer tutoring in SQL. The main goal is to develop a method for analyzing the student's SQL queries. The purposes: to form a general scheme and features of the method for analyzing student’s SQL queries based on the principles of technical diagnostics and methods of lexical, syntactic analysis of computer programs; to develop methods for parse tree construction; to create methods for comparing reference and real SQL queries according to their similarity rate; to demonstrate the function ability of the developed methods on specific examples. The methods used computer programs the automatic testing method, the computer programs lexical and syntactic analysis methods, the computer programs parsing trees construction methods, the objects diagnosing method based on comparison with a reference, the strings analysis methods, the method of q-grams. The following results were obtained: the student’s SQL queries analysis method was formed based on a system approach including automatic testing on real data, building query-parsing trees, comparison with a reference, and comprehensive determination of the queries similarity rate. The scientific novelty is the improvement of the method for the student's SQL query analysis during intelligent computer training in SQL query composition.