Grammar is known as a universal tool for the verbal representation. The research takes into account widely acknowledge grammatical features of legal texts. Scholars traditionally mention that they include the predominance of verb forms of the present tense, passive constructions, modal verbs to indicate the need and possibility, the absence of personal and demonstrative pronouns, the transmission of numerals, the use of Latinisms and cliches. It is also necessary to mention the lack of emotional coloring and complex syntax that helps to achieve the accuracy and unambiguity of legal formulations. Since the grammatical structure of the language reflects the system of logical connections with which the surrounding world is perceived and described, the grammatical structure for the translation process is largely determines the semantic content interpretation of the legal text in the course of translation. However, the generalized observations are not enough when it comes to multilingual lawyers and legal translators training. The required skills and abilities are subject to consistent training which requires systemization of challenges and stumbling block regarding concrete language layers and subsystems (lexis, grammar, stylistics, text level, genre aspect, etc.). Unfortunately, most of current educational aids on legal translation limit their material to the generalized assignments for the whole text translation, search for word combinations analogues, their replacement/matching, etc. The mentioned training toolkits do not focus on systematization of translation tools and prefer text-based approach. It might be reasonable as human speech is embedded in texts rather than in isolated phrases or words. However, it is necessary to overcome gaps in academic training practice and acknowledge that multilingual lawyers' training requires a concrete description of semantic and pragmatic zones within two working languages where there is no equivalence between two legal cultures/legislations, language structures. Thus the research goal is to consider major grammar challenges that should be born in mind in the course of legal translation as subject to legal translators and multilingual lawyers' training. The research findings focus on complex grammatical transformations. Respective techniques include word order changes, sentence structure changes, parts of speech and sentence member replacement, words addition and omission. The examples provided in the article confirm that the complexity of the process is often due to the fact that the mentioned techniques tend to be applied at the same time. The study outcomes can be used both in theory and practice of legal translation training. The relevance of the research concerns the development of legal translation as a theoretical discipline within interdisciplinary Language Studies, contributes to renewal of legal translation course syllabus and training toolkits, bridges the gaps between theory and practice.