Unstressed vowels in English play an important role in natural pronunciation. The importance of schwa, /э/ has been hardly recognized because its sound has no phonetic value. The purpose of this study applies the analysis of acoustic phonetics to compare the pronunciation of unstressed vowel of schwa, /э/ of Korean high school students to that of native speaker. One boy student and one girl student who speak English excellently in the high school are selected and a native speaker aged 30 from the United States who speaks American English is used as a standard. The approach of analysis is to measure 3 elements specifying the characteristics of unstressed vowels of schwa which are length, intensity and pitch, and to compare the data from the native speaker and that from the students. The differences or mistakes are analyzed to improve the pronunciation. The recording data is the eleven sentences of "Diagnostic Passage" quoted from "Manual of American English Pronunciation" (Prator & Robinett, 1985: 236). The speakers read the sentences and 35 syllables containing the unstressed vowels of schwa, /э/ are selected from the words in the unstressed sentences and analyzed by Praat, a phonetic analysis program. This study is based on "Duration and intensity as physical correlates of linguistic stress" (Fry, 1955: 765~768) to verify theory at the point of the unstressed vowels of schwa, /э/, that English speakers pronounce stressed syllables with fuller, clearer, louder and longer vowels than they do unstressed syllables. On the other hand, the vowels in unstressed syllables are often reduced and pronounced shorter, quieter and less clear (Pickett, 1980; Ladefoged, 2001). The results of analysis are as followed: Firstly, the length analysis of unstressed vowels indicated that the mean length of unstressed vowels pronounced by the native speaker was 55.4% of the mean length of stressed vowels whereas 88.7% of the boy student and 93.66% of the girl student respectively. Therefore, it is concluded that the students' pronunciations of unstressed vowels were almost indistinguishable from stressed vowels in length, as they also pronounced unstressed vowels much longer than the native speaker. The students also pronounced unstressed vowels in irregular and unstable manner. Secondly, the intensity analysis of unstressed vowels indicated that the mean intensity of unstressed vowels pronounced by the native speaker was 94.56% of the mean intensity of stressed vowels whereas 100.21% of the boy student and 97.77% of the girl student respectively. Similar to the first point, the students' pronunciations of unstressed vowels were almost indistinguishable from stressed vowels in intensity, and they pronounced unstressed vowels stronger than the native speaker. Thirdly, the comparison of pitch of the students with that of the native speaker indicated that the students' pitches were high on stressed vowels, but also on unstressed vowels were made at 7 of total 11 sentences on the analysis. Therefore, understanding of sentence stress and improvement of sentence rhythm were required. Fourthly, the correlation analysis between pitch and intensity showed that the pitch and the intensity of the students moved almost in parallel. On the other hand, the pitch and the intensity of the native speaker moved independently. The students mispronounced 8 of total 11 sentences on the experiment. It is concluded that the students were not sufficiently aware of unstressed vowels. Fifthly, the questionnaire analysis of the awareness of unstressed vowels indicated that the native speakers evaluated the pronunciation level of the students as 2 out of 5, equivalent to strongly unsatisfied. It means that the most of the students were insufficient in pronouncing unstressed vowels. In fact, 89.5%(17 of 19) of the students answered that they knew almost nothing about the unstressed vowels. Only 15.8% of the students had the correct answers for the evaluation test of unstressed vowels results. Consequently, the recognition of unstressed vowels of the students was significantly low. In conclusion, this analysis is in the category of a pilot study of the unstressed vowels. The pronunciation training should be performed on the basis of the acoustic theory and the physical measuring because there are a lot of pronunciation errors on the unstressed vowels of the students. This is a meaningful acoustic analysis to measure the pronunciation of unstressed vowels. The scientific approach to identify the mistakes of the students in English pronunciation should be performed.