Background: Allocation of trainee therapist cases is often performed based on intuition and clinical circumstances, with lack of empirical evidence on the role of severity of presenting problem. This has the potential to be anxiety-provoking for supervisors, trainees and service users themselves. Aims: To determine how therapist experience interacts with symptom severity in predicting client outcomes. Method: An intention-to-treat analysis of annual outcome data for primary and secondary care clients seen by a specialist anxiety disorders service. 196 clients were stratified into mild, moderate and baseline severe symptoms of anxiety (GAD-7) and depression (PHQ-9). We measured percentage change on these measures, as well as number of sessions and therapy dropout. We also examined rates of reliable and clinically significant change on disorder-specific measures. We hypothesized that qualified therapists would achieve better outcomes than trainees, particularly for severe presentations. Results: Overall, outcomes were comparable between trainee and qualified therapists on all measures, and trainees additionally utilized fewer therapy sessions. There was however an interaction between anxiety severity (GAD-7) and therapist group, such that severely anxious clients achieved greater symptom improvement with qualified as compared to trainee therapists. Further, for trainee but not qualified therapists, baseline anxiety was negatively associated with rate of reliable and clinically significant change on disorder-specific measures. Conclusions: These findings indicate generally favourable outcomes for trainee therapists delivering manualized treatments for anxiety disorders. They additionally suggest that trainee therapists may benefit from additional support when working with clients that present with severe anxiety.