BACKGROUND: Sound errors are characteristic of acquired apraxia of speech (AOS) and are frequently the focus of treatment. One treatment for AOS, Sound Production Treatment (SPT), has been shown to facilitate improved sound production in trained sounds and untrained exemplars of those sounds (Wambaugh, Kalinyak-Fliszar, West, & Doyle, 1998a). Although the effects of SPT are relatively well understood when the treatment has been applied sequentially to single sounds in words, little is known about its application to multiple sounds. Additionally, the stimulus generalisation effects of SPT have not been well specified. AIMS: This investigation was designed to further investigate the acquisition and stimulus generalisation effects of SPT for AOS. Treatment application was modified from previous investigations to allow for application with multiple sounds and in a different treatment context. METHODS AND PROCEDURES: A multiple baseline design across behaviours and contexts was used to assess the effects of treatment with a speaker with chronic AOS and aphasia. Treatment was initially applied within the context of words elicited through repetition. In order to assess stimulus generalisation, the nine consonants of interest were elicited in words through sentence completion. Additionally, the target sounds were elicited in a different word position (i.e., word-final) from that utilised in treatment (i.e., word-initial). OUTCOMES AND RESULTS: Treatment resulted in increased correct productions of the target sounds in trained contexts. Generalisation to the different stimulus contexts was limited and varied across sounds. Treatment was extended to the sentence completion context and additional treatment gains were observed. CONCLUSIONS: The application of SPT to multiple sounds and to an additional treatment context appears to have promise in the treatment of AOS. Additional replications are required for external validity.