The article is aimed at examining the potential significance of frequency and compatibility of the forms of the lemma eye, represented in the corpus, for the actual semantics of the English idioms with the corresponding key components. The component composition of the inner form of the idiom and its correlation with the corresponding lemmas in the concordance are in the focus of attention. The semantics and structure of English idioms have never been examined from such an aspect before. The research reveals that the verbs characterized by their frequent combination with the lemma eye are also represented in the idioms with the components eye/eyes and are definitely important for the formation of the semantics of these idioms. Some other frequently encountered combining features of the lemma eye, for example, its combination with attributes, as well as the prevalence of its plural form in the corpus, are found to be not that significant for the actual semantics of the idioms. The research is based on the postulate that both idioms and corpus reflect the perception of eyes by the native speakers’ community. This assumption makes it possible to compare it with the conceptual representation of eyes in the corresponding idioms. The irregularity of semantic and conceptual construction of idioms and originality of the associative connections involved are confirmed by the quantitative indicators. As a result, a new line of research, which consists in the study of the semantic and conceptual structure of idioms against the background of corpus data, is proposed.