The article was prepared within the framework of the scientific project No. 16-03-50161, supported by the Russian Foundation for Basic Research. Realism and formalism are described in scholarly literature as two styles of judicial interpretation. This is not free from logical error of incorrect inference from the assumed grounds. The paper examines the realist approach to the methodology of judicial interpretation in comparison with the formalistic style of judicial interpretation, and formulates the critical arguments that discover both theoretical and practical difficulties of methodological position of that approach, which is gaining the leading position in the doctrine, primarily foreign, and is largely determining the policy of the judicial interpretation in the institutions of “high justice”. The realistic style of judicial interpretation has no clear connection with any particular type of legal philosophy and can not be reserved solely for representatives of legal realism (American and Continental). The realistic approach to judicial interpretation is revealed by the following six theses: 1) voluntarism in the interpretation; 2) semantic indeterminacy of the text; 3) freedom of interpretation; 4) logical skepticism; 5) decisionism; 6) mechanical application of law. Together they characterize, respectively, (1) nature, (2) object (subject) of interpretation, and (3) methods of using it, (4) results of interpretation and (5) its validity and effectiveness in legal order, (6) relation of realistic approach to the model of subsumption as attributes of formalistic style of judicial interpretation. Each thesis can exist in its “hard” and “soft” form, resulting in a typology of judicial realism in radical and moderate variants, despite the fact that common features of two versions of the realist approach are: 1) the interpretation of legal norms as teleological judgments; 2) admission of metalegal arguments to the basis and the structure of judicial reasoning. The comparative analysis of moderate realistic and radical-realistic styles of judicial interpretation reveals the absence of fundamental differences between formalistic and moderate realist positions, which are regarded as complementary, rather than mutually exclusive. In relation to radical realist approach the author formulates critical arguments about: 1) redundancy of the thesis of semantic uncertainty in relation to the concept of semantic-normative voluntarism; 2) infinite regress of acts of interpretation and the impossibility of norms; 3) incompatibility of the denial of the general rules and the approval of the validity of court decisions; 4) self-reference of the acts of interpretation and powers of the subjects of interpretation; 5) the impossibility of mechanical application of law.