the present study applied a benchmarking strategy to evaluate the outcomes of youth (6 -15 years) with anxiety disorders treated at 'systems of care' children'smental health services (soc cmhs). there were three stages of analysis. the first used meta-analytic technique to aggregate results of 17 randomised controlled trials of treatments of anxiety in youth. from these studies, benchmarks were established for two different outcome criteria: pre-post effect sizes and the proportion of youth evidencing 'clinically significant improvement'. two subsets from the soc cmhs data were considered. the first was comprised of youth who were selected on the basis of a combination of child behavior checklist profile and dsm diagnosis or presenting problem. the second was comprised of youth selected primarily on the basis of clinician-generated dsm diagnosis. neither subset attained levels of improvement commensurate with treatment efficacy benchmarks. only one subset (selected partly on the basis of child behavior checklist profile) achieved results significantly better than natural remission and this was only for one natural history benchmark (pre-post effect size). the third stage of analysis examined factors associated with reliable improvement and treatment response. results indicated that the relatively poor response of youth from the soc cmhs agencies could not be explained by the socio-psychological characteristics of this group. avenues for future research are suggested, including extension of benchmarking strategies in children's mental health and improved understanding of predictors of treatment response.