Aims and method We examined learning outcomes, practice impacts and implementation processes for a training intervention in diagnostic skills delivered to multidisciplinary child and adolescent mental health service practitioners (n = 63). Results Training was viewed positively by most participants and associated with significant increases in practitioner self-efficacy, with the effect sustained at 8-month follow-up. A comparative audit before and after training indicated that clinicians were significantly more likely to assign an Axis I diagnosis following the training intervention. However, absolute rates of Axis I classification remained relatively low (< 40%) both before and after training. Practitioners were moderately successful at following through on personal plans for implementing new learning; inconsistent support for implementation was provided within teams. Clinical implications A brief training workshop may have limited effects in changing practitioners' behaviour so that diagnoses are made more promptly and appropriately recorded. Future workforce development initiatives should consider more comprehensive and diversified strategies, including targeted post-training support, if increased self-efficacy following training is to be translated into sustained changes in diagnostic practice.