Abstract: The matter under research is diagnostic capabilities of psychological tests. Despite frequent use, this term does not have a clear definition and can be interpreted in a different way in researches. The authors of the article examine such aspects of the topic as the use of the term 'diagnostic capabilities' in Russian empirical reserach, the structure of the term 'diagnostic capabilities' of psychologists tests, problems caused by the use of unreliable tests in psychological practice, reasons and solutions. The researchers have also carried out a practical analysis of the term 'diagnostic capabilities' as it has been used in recent empirical researches and have analyzed objective reasons of why unreliable psychological tests are commonly used. The main conclusions of the research are the following: 1. there is the 'theory-and-practice' dialectics in academic psychology: borders of 'diagnostic capabilities' are defined by the theoretical model of psychic phenomenon which, in its turn, is based on actual empirical data. 2. The definition of 'diagnostic capabilities' should include measurable effects and diagnostic attributes; applicability for particular research objectives and groups of respondents, standard samples or reference patterns that can be used in the process of interpretation, reliability, reasonability and validity. The term 'diagnostic capabilities' is identical to the term 'psychometric properties'. 3. Objective reasons for such common use of unreliable psychologists tests include: the use of out-of-date tests; availability of such tests; the fact that tests are used by other specialists but not psychologists; and use of computer interpretation of tests. 4. The authors describe ways to overcome this problem, i.e. to involve students in psychologists researches carried out by university departments; to develop the mechanism of distribution of psychologists tests; awareness of harmful effect of online diagnostics; increasing availability of online counselling; and encouraging better understanding of methodological bases of particular tests.