Jeon Ji-eun. 2014. A Study on Key Words Analysis according to Social Variables of Gender and Age using Sejong Korean Spoken corpus. The Sociolinguistic Journal of Korea 22(1). This study shows how a key words and key clusters analysis offers an empirical data of the significant characteristics in social variables of gender and age based on Sejong spoken corpus. “Key words” is a term for statistically significant lexical items which are most frequent in a given corpus compared with other corpus as a reference (Scott 1999). We also extend the keyness concept to word combinations, clusters which are found repeatedly together in each others' company, in sequence. We use the KeyWords facility in WordSmith tools (Scott 2012) which has been used in several studies as means for describing the characteristics of social variables. We extract (1) the male and female key words/key clusters in public and private speech, (2) the 20s and over 30s key words/key clusters in public and private speech. Therefore, we found that these empirical investigations verify the lexical saliency of utterance by gender and age. The gender key words show more prominent features than age key words. Moreover, The key words analysis is more efficient than key clusters analysis in sociolinguistic studies. This study is illustrative of the potentiality of the corpus-based research on social differentiation in the use of language.