In spite of numerous theoretical studies on sentential negation in Korean (an V for short form negation and V-ci anh for long form negation) in the literature, much of the discussion has focused on its abstract features, relying on the researcher`s intuition or introspective judgment of limited sources. This paper presents an empirical observation by investigating the actual distribu- tion of sentential negation in corpus, based on the Sejong Spoken Corpus. Two analyses of sentential negation were done: frequency analysis and compatibility with predicate types. The frequency analysis revealed that long form negation was used more often than short form negation. The analysis in terms of compatibility with predicate types showed that long form negation appeared more frequently with Sino-Korean predicates and with adjectival predicates than short form negation. The findings were discussed in relation to register variation and discourse strategies.