Abstract In comparison to the contributions that address interlingual phraseological equivalence, the number of papers about the topic of phraseological false friends is relatively low. This is probably explained by the fact that this is a marginal phenomenon from a quantitative point of view. Nonetheless, there are relevant contributions in the field of foreign German Studies. The aim of this article is, on the one hand, to discuss theoretical questions about tertium comparationis and terminology and, on the other hand, to develop a classification of the types of potential phraseological false friends. This paper must therefore be understood as a preliminary stage for applied work in the areas of foreign language teaching, (not only) bilingual lexicography and translation theory and practice. Based on a comprehensive empirical basis, the present article studies nominal, adjectival and especially verbal idioms. For the collection of both corpora and the codification of the lemmas various methods were used: the consultation of specialized dictionaries, the use of the author’s own foreign and native language competence, the search of context examples from the databases Corpus de referencia del español actual (CREA) and Deutsches Referenzkorpus (DeReKo), and the consultation of informants. Through this methodological approach, the article tackles the levels of the language system and also of the language usage. In this respect, and due to the problem of the so-called broad or complex meaning of idioms, difficulties associated with the analysis of lexicographic definitions are of particular relevance. Depending on the lexicographic sources, differences in questions such as the complexity of the definitions or the number of sememes are noticeable. On the level of meaning, componential analysis represents the theoretical framework. Thus, the semantic structures of the idioms to be compared are analyzed in order to discover to what extent the whole meaning (sememe) or some of the minimal semantic features (semes) of the units are qualitatively different or in unequal numbers. Differences can be found both at the level of monosemic units – where either (i.1) the sememes of the units to be compared is basically different, or (i.2) one or several semes of the units to be compared differ – and also at the level of polysemic units, if the form of the idiom of one language with several sememes finds an identical or similar form in the other language but the latter does not have the same sememes. The semantic analysis performed is thus the basis for the determination of the different types of semantic interferences, which can lead to communicative misinterpretations to various degrees. On the level of expression, the analysis is based on a structural-cognitive hypothesis that postulates both a figurative and a logical-abstract formal identity. The formal similarity between idioms (and other types of phraseologisms) in different languages is not rooted in the phonetic and graphical (quasi-)identity of the units to be compared, but in the identity or in the somewhat similar height of the phrase image. Beyond such a concept of lexical-figurative identity, we find a broader conception of formal equality, which is understood not only as a structural lexical-figurative identity, but also as a likeness or identity of the logical-hidden scheme beyond the image; this scheme can initiate the same idiomatic inference procedures, even if the idioms diverge figurative-lexically. A meticulous interlinguistic analysis of the phraseological false friends is only possible by means of a clear distinction between both levels of meaning and expression, which must be reflected terminologically. In this sense, we refer to the proposal of B. Wotjak, which uses the term Kongruenz to denote cases of equality of linguistic forms, as a counterpart to the term Äquivalenz on the content level, i.e. the equality on the level of meaning. The term interlingual homonymy, on the other hand, is avoided because of the special nature of formal equality between idioms.