In this paper, we compute entropies and average mutual informations for a 'choseong,' a 'jungseong,' and a 'jongseong' in multi-syllable Korean words. In these computations, we consider the property that each Korean syllable is divided into a 'choseong,' a 'jungseong,' and a 'jongseong' which are a consonant, a vowel, and a consonant, respectively, without exception. We compute entropies and average mutual informations for a 'choseong,' a 'jungseong,' and a 'jongseong' from the cumulative frequency of Korean syllables with respect to word length and syllable location in a word. We compute average mutual informations for a syllable and for a 'choseong,' a 'jungseong,' and a 'jongseong' between the two adjacent syllables in a word. The correlation between the two adjacent syllables in a word is not large on the average.