Abstract:  Based on Peter Skehan′s ′input-central processing-output′ framework, the influence of audio input and audio-visual input on EFL learners′ oral fluency was investigated by using both quantitative and qualitative methods Non-English majors were chosen as subjects At the first stage, subjects received audio input, prepared for two minutes and then orally produced one-minute personal statement At the second stage, subjects received audio-visual input and then produced another one-minute personal statement Speech samples were collected and analyzed by Praat and SPSS Findings show that there are statistically significant differences (p<005) in the number of pauses and mean length of runs, with obviously fewer pauses and longer mean length of runs after receiving audio-visual input However, the other three indices, mean length of pauses, mean number of repairs and syllables, are not significantly different Therefore, we may tentatively conclude that EFL learners produce more fluent oral production when they receive audio-visual input

Key words: oral fluency, audio input, audio-visual input