Globalisation and modern information technologies cause the spread of cognates in languages of the world. Displaying a similar form and content but exhibiting different patterns of usage, cognates seem to become a pitfall for foreign language users and translators. A possible solution to the problem is the analysis of cognate usage based on comparable corpora which offer a reliable source of data for this kind of analysis. The paper deals with one lexeme in English, apathy, and its equivalent in Lithuanian apatija and their concordances compiled from the Bank of English and the Corpus of Contemporary Lithuanian. The data are analysed and classified into types of word combinations such as subjective, predicative, objective, attributive, and adverbial. Enumerative noun chains and afterwords are compared between the two languages as well. Similarities and differences in the apathyapatija usage patterns are disclosed and some generalisations drawn. The article suggests that the English noun apathy is more common and universal in nature and it is used in a greater variety of different contexts and situations than is the case with its Lithuanian counterpart apatija. Thererore, it is useful for foreign language learners, users and translators to resort to corpora due to the fact that monolingual and bilingual dictionaries cannot be treated as reliable sources of the correct usage of collocations.