Background: There has been little robust evaluation of the outcome of speech and language therapy (SLT) intervention for post-stroke dysarthria. Non-speech oro-motor exercises (NSOMExs) are a common component of dysarthria intervention. A feasibility study was designed and executed, with participants randomized into two groups, in one of which NSOMExs were a component of the intervention programme. Aims: To examine (1) operational feasibility of the programme; (2) participants' views of the programme; and (3) speech intelligibility, communication effectiveness and tongue and lip movement at four points (A1 and A2 before, and A3 and A4 after intervention). Methods and Procedures: Thirty-nine participants were randomized into Group A (n = 20) and Group B (n = 19). Groups were equivalent at enrolment in demographic variables and A1 measures. Intervention was behavioural, delivered in eight home-based SLT sessions, and included practise of individually appropriate words, sentences and conversation, and for Group B also NSOMExs. Between-session practice was recorded in a diary. Data on speech intelligibility, effectiveness of communication in conversation, self-rated situational communication effectiveness, and tongue and lip movement were collected at 8-week intervals, twice before and twice after intervention. Anonymous evaluation (AE) questionnaires were completed. Outcome and Results: The recruited number was 20% below the target of 50. Thirty-six participants completed the intervention and 32 were followed through to A4. The programme was delivered to protocol and fidelity was verified. Thirty-four AEs were returned. These showed high satisfaction with the programme and its outcome. According to diary records from 32 participants, 59% carried out at least the recommended practice amount. Outcome measure performance across the four assessment points did not indicate any group effect. For the whole sample both externally rated and self-rated communication effectiveness measures showed statistically significant gains across the intervention period (A2/A3), which were maintained for 2 months after intervention (A2/A4). Non-intervention period changes (A1/A2 and A3/A4) were not present. There were no intervention-related gains in tongue and lip movement or speech intelligibility, but the latter is likely to be attributable to a ceiling effect on scores. Conclusions and Implications: The results indicate positive outcomes associated with a short period of behavioural SLT intervention in the post-stroke dysarthria population. The inclusion of NSOMExs, delivered in accordance with standard clinical practice, did not appear to influence outcomes. The results must be viewed in relation to the nature of feasibility study and provide a foundation for suitably powered trials.