The voicing contrast of word final obstruents in English is reliably perceived in the duration of the preceding vowel. The vowel preceding a word final voiced obstruent is longer than the vowel preceding a word final voiceless obstruent. In this study, monosyllabic words with final obstruents, /p-b/, /t-d,/ /k-g/, /f-v/, /θ-ð/, and /s-z/ were collected in stressed and unstressed conditions. The measurement of segment duration showed that native English speakers produced significant lengthening of the vowels preceding word final voiced consonants. The exception was unstressed words with non-sibilant fricatives (/f-v/, /θ-ð/). For the Korean learners of English, vowels preceding alveolar stops (/t-d/) and alveolar fricatives (/s-z/) have significant durational differences regardless of the placement of stress on the words. However, the speech of Korean learners of English presented inconsistency of the vowel duration differences and the duration change was smaller in magnitude compared to the data of native English speakers. The study results were discussed in relation to the consonant inventory and vowel durational pattern of English and Korean as well as their implication for teaching L2 English pronunciation.