From 1995 onwards, a child and adolescent mental health service (CAMHS) applied Senge's learning organisation model. This review compared service performance with that of peer services 5 years later and explored whether any differences were associated with the application of this model. The comparison methodology used quantitative analysis of external data from the Department of Human Services, together with qualitative analysis of material including interviews with CAMHS directors and service managers. Results showed high evaluation activity and high quality, efficiency and efficacy of care compared with other services. Several restraints to the optimal application of the model were identified, including inadequate training of new managers, service overload, major external organisational change and limited investment in information systems. Other outcomes are discussed.