Objective: To compare the effects of different types of clinician feedback (auditory knowledge of performance, KP, kinaesthetic KP and knowledge of results, KR) on the production of a voice task during acquisition. KP conditions directed attention to auditory or kinaesthetic cues. KR feedback was provided as to how close the speaker's production was to a model. Patients and Methods: A factorial multivariate 3 × 8 design was used. Thirty-six females were randomly assigned to 3 different feedback groups and trained in the production of a vocal siren. The production of sirens was measured by variants of vocal intensity, cepstral peak prominence, phonation time and pitch phase compliance. Results: All groups showed significant improvement over time in measures of vocal intensity, timing of pitch change and phonation time. The KP group (auditory and kinaesthetic combined) showed significantly more improvement over time in vocal intensity than the KR group. The kinaesthetic KP group maintained better control of vocal clarity across attempts while reducing vocal intensity than the auditory KP group. Conclusion: Optimal feedback paradigms during acquisition may be different depending on the parameters of voicing being trained. Learners may benefit from KP feedback to control intensity, and kinaesthetic cues to control vocal clarity.