Background: There is an increasing interest in measuring health care outcomes in mental health services for children as well as adults.

Aims: We examined the sensitivity of the Paddington Complexity Scale (PCS) and the Health of the Nation Outcome Scales for Children and Adolescents (HoNOSCA) in describing the intakes of child and adolescent mental health clinics.

Method: We carried out a prospective study of two out-patient units and one day patient unit, by means of questionnaires administered to clinicians, parents and children.

Results: Clinician-rated PCS and HoNOSCAs were obtained for 248 new attenders. Both proved sensitive to intake differences between clinics. There were correlations of moderate intensity (r = 0.6) between total PCS and HoNOSCA. There were weaker associations between these and parent/child rated measures of behaviour (r = 0.4 and 0.3), quality of life and self-esteem (r = 0.3 or less).

Conclusions: Both PCS and HoNOSCA are useful for describing clinical profiles of children and adolescents receiving mental health services.