BACKGROUND: In clinical practice, it is common for speech and language therapists to discontinue dysarthria treatment when a plateau in spontaneous recovery is reached. However, there is some evidence from single-case research studies that people with chronic dysarthria can improve their speech with intervention several years after the onset of the disorder. If this is so, it can be argued that speech treatment should still be an option available to them. Motor learning principles suggest that in order to be effective, treatment needs to involve frequent, consistent practice with feedback. AIMS: To compare the effects of both traditional treatment for clients with longstanding, stable dysarthria with computerized speech therapy, which allows independent practice with feedback. DEVELOPMENT OF THE COMPUTER PROGRAM: A new computer program was developed through a European Union-funded, quality of life and management of human resources programme. The program was designed for use with people with speech disorders and was based on motor learning principles. Speech recognition technology was exploited to provide feedback on practice attempts. METHODS & PROCEDURES: Seven clients with longstanding, stable dysarthria completed a within-participant ABAC/ACAB case series design to compare 6-week blocks of weekly traditional treatment with 6-week blocks of computer-based treatment over a 6-month period. OUTCOMES & RESULTS: The findings suggest that the participants improved their speech with treatment. The computerized therapy was as effective as traditional treatment. The participants spent more time (37%) with the computerized therapy than with traditional practice techniques. CONCLUSIONS: The participants with longstanding dysarthria in this study made improvements to their speech with treatment using both traditional and computer-based techniques. Computers could provide a method of delivering an effective dysarthria treatment service without placing high demands on therapy resources.