Cognitive Behaviour Therapy (CBT) has the strongest preliminary support for treatment of depression and anxiety following traumatic brain injury (TBI). TBI associated cognitive impairments may pose an obstacle to development of a strong working alliance, on which therapeutic gains depend. The current study examined the association of demographic (i.e., gender, age at study entry, years of education and premorbid IQ) and injury-related (i.e., years since injury, post-trauma amnesia duration, memory and executive functioning test performance) variables with alliance in CBT adapted for TBI (CBT-ABI). The audio-recordings of 177 CBT-ABI sessions from 31 participants were assessed with an observer version of the Working Alliance Inventory at nine time-points. Multi-level mixed model regressions showed that participants and therapists maintained a relatively strong alliance across all sessions. Pre-intervention symptom severity was considered as a confounder variable and was found to have no statistically significant influence on the models. None of the demographic variables were significantly associated with alliance scores. More years since injury was associated with a stronger alliance. These findings demonstrate that TBI associated cognitive impairments do not necessarily pose an obstacle to development and maintenance of a strong working alliance, which is more likely to develop with more time post-brain injury.