Primary care patients frequently present with anxiety with prevalence ratios up to 30%. Brief cognitive–behavioural therapy (CBT) has been shown in meta-analytic studies to have a strong effect size in the treatment of anxiety. However, in surveys of anxious primary care patients, nearly 80% indicated that they had not received CBT. In 2010, a model of CBT (Coordinated Anxiety Learning and Management (CALM)) adapted to primary care for adult anxiety was published based on results of a randomised controlled trial. This project aimed to integrate an adaptation of CALM into one primary care practice, using results from the published research as a benchmark with the secondary intent to spread a successful model to other practices. A quality improvement approach was used to translate the CALM model of CBT for anxiety into one primary care clinic. Plan-Do-Study-Act steps are highlighted as important steps towards our goal of comparing our outcomes with benchmarks from original research. Patients with anxiety as measured by a score of 10 or higher on the Generalized Anxiety Disorder 7 item scale (GAD-7) were offered CBT as delivered by licensed social workers with support by a PhD psychologist. Outcomes were tracked and entered into an electronic registry, which became a critical tool upon which to adapt and improve our delivery of psychotherapy to our patient population. Challenges and adaptations to the model are discussed. Our 6-month response rates on the GAD-7 were 51%, which was comparable with that of the original research (57%). Quality improvement methods were critical in discovering which adaptations were needed before spread. Among these, embedding a process of measurement and data entry and ongoing feedback to patients and therapists using this data are critical step towards sustaining and improving the delivery of CBT in primary care.