Audio-assisted Repeated Reading (RR) is a method which scaffolds L2 readers to build fluency. This study focused on 27 U.S. college-level learners of Japanese comprising three ability groups from mid-beginner to low-intermediate. It aimed to explore whether a modest semester-long RR program facilitates learners' fluency growth, and see how the learners perceive the method itself using a questionnaire. We found beneficial effects of audio-supported RR in terms of reading rate growth while learners maintained good comprehension. The study showed a significant practice effect, in which learners increased their reading rates while re-reading the same passage. Further, learners in one of the three groups read significantly faster with a new, unpracticed passages representing transfer of the practice effect. Additionally, questionnaire responses demonstrated that learners perceived beneficial effects from RR. Finally, we make a proposal on measuring learners' reading comprehension in fluency intervention programs such as RR. We proposed that measurement of learners' reading comprehension have two different purposes, pedagogical purposes and research purposes. Thus the reading comprehension measures would have different designs and cut scores.