In order to provide “a common basis for the elaboration of language syllabuses, curriculum guidelines, examinations, textbooks, etc. across Europe”, The Common European Framework for Languages (CEFR) was published in 2001 by the Council of Europe. It has affected the way languages are taught, learnt and assessed and also how foreign language proficiency levels are defined all around the world. The CEFR adopts an intercultural approach to foreign language, and the main purpose is to protect cultural diversity and to give importance to cultural activities rather than being a part of foreign language education. For this reason, culture is at the very core of the CEFR. In 2018 and 2020, two Companion Volumes were published to complement the CEFR. The present paper offers a comparative corpus analysis of these three texts focusing on the occurrences of culture-related items using n-gram tool of Sketch Engine (Lexical Computing, n. d.), which creates frequency lists of sequences of tokens. Based on the findings, it is suggested according to the CEFR that rather than focusing on the national culture of the native speakers of the target language, foreign language education should focus more on the “new culture” formed by the encounters of people coming from different cultures.