This article details the transcription process we developed to handle bimodal and multilingual interview data collected during our research into the teacher language awareness (TLA) of two high-quality teachers in a Swedish bilingual school for the deaf. Both teachers used a combination of spoken Swedish, spoken English, and Teckenspråk (Sweden's signed language) during our interviews. Because the interviews were being conducted in three languages and two modalities, the transcription process was particularly complex. We argue that accurate transcription in such linguistically complex contexts is possible but that it is important that the research community develop principles and procedures for transcribing multilingual and bimodal interview data in a way that honors participants' intended messages as faithfully as possible.