Background & Aims: We evaluated a structured intervention programme aimed at preparing adolescents with developmental language disorders for job interviews. Our primary outcome measures included change in ratings of verbal and non-verbal social communication behaviours evident during mock interviews. Methods & Procedures: In study 1, 12 participants, aged 17-19 years, from a specialist sixth-form college completed the intervention and two mock interviews, one pre- and one post-intervention. In study 2, 34 participants, aged 17-19 years, completed a modified intervention programme and three mock interviews, one at baseline (included to control for possible practise effects), one pre- and one post-intervention. In both studies, interviews were video recorded and social communication behaviours were coded by independent assessors blind to interview time, participant diagnosis and therapy content. A repeated-measures design was employed to measure change in communication behaviours. Outcomes & Results: In study 1, a significant increase in the number of "positive" verbal and non-verbal social communication behaviours was observed from pre- to post-intervention. However, there was no significant change in the number of "negative" behaviours (i.e., fidgeting, irrelevant remarks). In study 2, there were no significant changes in verbal behaviours, but significant group differences (though wide individual variation) in both positive and negative non-verbal social communication behaviours. Conclusions & Implications: Our findings suggest that training specific social communication skills that are important for interview success, and consistently reinforcing those behaviours during therapy practice, can increase the use of those skills in an interview setting, though in this heterogeneous population there was considerable variation in therapy outcome. The skills of the interviewer were identified as a potential source of variation in outcome, and a target for future research and practice.