Background: In recent years, analyses of cleft palate speech based on phonetic transcriptions have become common. However, the results vary considerably among different studies. It cannot be excluded that differences in assessment methodology, including the recording medium, influence the results. Aims: To compare phonetic transcriptions from audio and audio/video recordings of cleft palate speech by means of outcomes of per cent correct consonants (PCC) and differences in consonant transcriptions. Methods & Procedures: Thirty-two 3-year-olds born with cleft palate were audio and audio/video recorded while performing a single-word test by picture naming. The recordings were transcribed according to the International Phonetic Alphabet. The transcriptions from the audio and the audio/video recordings were analysed with regard to PCC, differences in the use of phonetic consonant symbols and the use of diacritics. Outcomes & Results: PCC was higher when the calculation was based on transcriptions from audio recordings than when based on audio/video recordings. No such differences were seen when age-appropriate consonant processes were scored as correct. Consonants were not excluded as often in the transcriptions from audio/video recordings as they were in the transcriptions from audio recordings, and more target consonants were transcribed as dental/alveolar in the transcriptions from audio/video recordings and palatal/velar/uvular in the transcriptions from audio recordings than vice versa. Further, interdental articulation, linguolabial articulation and audible nasal air leakage were more common in the transcriptions from audio/video recordings than in those from audio recordings. All these differences were statistically significant. Conclusions & Implications: Phonetic transcription is influenced to some extent by visual cues. However, as long as age-appropriate articulatory and phonological simplification processes are scored as correct when evaluating consonant production in the speech of young children born with cleft palate using phonetic transcription, the recording medium does not seem to affect the results.