Existing literature broadly demonstrates that a therapist’s tone of voice affects interpersonal and therapy outcomes, and qualities of human voice tone (pitch, volume, rhythm, speed and intensity) influence human interaction and attachment. However, little empirical and systematic evidence surrounds the use of voice in Guided Imagery and Music (GIM) and other psychotherapeutic practices. This exploratory study investigates the use of acoustic analysis to report on ways in which therapists’ and clients’ voices attune with each other and the music in GIM practice. Seven volunteer university student participants each attended a single GIM session with a qualified GIM therapist. Audio recordings of the session were made using a multi-channel recording device and 61 excerpts of the recordings were then analysed using the MIRToolbox for Matlab. It was found that acoustic features of the client’s voice predicted acoustic features of the therapist’s voice more commonly than the other way round (therapist predicting client). Furthermore, both such predictions occurred more commonly than they did for acoustic features of the music predicting acoustic features of either the therapist or client. These results suggest the client’s voice influences the therapist’s voice more than the therapist’s voice influences the client’s, and the music is the least influential. This study developed a reliable methodology for data collection and analysis to examine interactions between the client and therapist voices and music in GIM practice and provides further insights for clinical practice and training, enhancing sensitivity to vocal interactions and attunement.