Given the rapid advances in technology over the past decade, this thesis examines the potential for automatic speech recognition (ASR) technology to expedite the process of objective analysis of speech, particularly for lexical stress patterns in childhood apraxia of speech. This dissertation also investigates the potential for mobile technology to bridge the gap between current service delivery models in Australia and best practice treatment intensity for CAS. To address these two broad aims, this thesis describes three main projects. The first is a systematic literature review summarising the development, implementation and accuracy of automatic speech analysis tools when applied to evaluation and modification of children’s speech production skills. Guided by the results of the systematic review, the second project presents data on the accuracy and clinical utility of a custom-designed lexical stress classification tool, designed as part of a multi-component speech analysis system for a mobile therapy application, Tabby Talks, for use with children with CAS. The third project is a randomised control trial exploring the effect of different types of feedback on response to intervention for children with CAS. The intervention was designed to specifically explore the feasibility and effectiveness of using an app equipped with ASR technology to provide feedback on speech production accuracy during home practice sessions, simulating the common service delivery model in Australia. The thesis concludes with a discussion of future directions for technology-based speech assessment and intensive speech production practice, guidelines for future development of therapy tools that include more game-based practice activities and the contexts in which children can be transferred from predominantly clinician-delivered augmented feedback to ASR-delivered right/wrong feedback and continue to make optimal gains in acquisition and retention of speech production targets.