PURPOSE: In a pilot randomized controlled trial, we investigated the effectiveness of a 12-weekly anxiety treatment programme adapted for individuals with moderate-severe TBI, based on cognitive behaviour therapy (CBT) and Motivational Interviewing (MI). The current study explored the variables associated with treatment response and group differences in change expectancy and working alliance. METHODS: Twenty-seven participants recruited from a brain injury rehabilitation hospital were randomly assigned to MI + CBT, non-directive counselling (NDC) + CBT and treatment-as-usual and assessors were blinded to treatment conditions. Correlation and multiple regression were used to examine the association between reduction in anxiety ratings and a number of clinical, injury and cognitive variables. Random effects regression was used to examine group difference in changes in working alliance and expectancy. RESULTS: There was a trend suggesting that greater injury severity may be predictive of poorer response to CBT and injury severity was significantly related to memory functioning. Participants receiving MI pre-treatment showed a greater increase in change expectancy at the end of CBT and at follow-up, but not at the end of MI. CONCLUSIONS: There is a need to further investigate the effectiveness of treatment for individuals with different injury severity and to explore the relationship between change expectancy and treatment outcome.