This paper aims to examine and evaluate the preceding literature on advertising language, and suggest a new methodology of analysing advertising language in terms of Corpus Linguistics. In section 2, the characteristics of preceding literatures are reviewed according to 4 distinctive linguistic domains-phonology, morphology/ lexicology, syntax and stylistics. After that, some points of dispute are raised with their alternatives. In section 4, using TV CF texts and Sejong Corpus, I tried to reveal a specific property of advertising language, or colloquial style, which has been argued by some preceding articles and books. According to the stochastic analysis based on the current data and Sejong Corpus, it’s hard to say that advertising language is a typical colloquial style. The stochastic result leads to the conclusion that advertising language would be placed between colloquial corpus and written corpus in terms of colloquiality.