Research shows that anxiety disorders are common and problematic in children. Treatment studies demonstrate that cognitive-behavioural interventions for children can successfully minimise these problems. Further, when implemented as early intervention or prevention programs, these interventions can prevent the onset of anxiety problems in 7 to 14 year olds and reduce existing levels of anxiety. This type of preventive approach has enormous potential for improving community mental health in a low cost model of service delivery. Yet, to develop this as a viable service model, these programs need to be evaluated under 'real world' conditions as opposed to specialist university clinical teams. In this research, the long-term effectiveness of an ecological model of the prevention of anxiety disorders for children was conducted. In the first study, teachers (N = 17) were trained intensively in the principles of anxiety and the FRIENDS anxiety prevention program (Barrett, Lowry-Webster & Holmes, 1999). Measures were taken of the PROXIMAL effects of training on the knowledge and self-efficacy of participating teachers at two points in time. Compared to teachers in a control group (N= 17) and a group comprised of psychologists regarded as experts in the FRIENDS anxiety program (N= 22), trained teachers demonstrated significant increases in the levels of knowledge and self-efficacy at time two. These increases approached levels exhibited by experts in terms of knowledge, and were not significantly different from experts in their levels of self-efficacy to implement the FRIENDS program following training. This study also explored the quality or fidelity of program implementation by these trained teachers (INTERMEDIATE EFFECTS). Results demonstrated that trained teachers implemented the program with high levels of integrity in accordance with the FRIENDS intervention manual. The second study sought to investigate the outcomes for participating children in terms of actually preventing and reducing existing levels of anxiety. Also of interest was the impact this intervention had on levels of depression. Participants were 594 children aged 10-13 years from 7 schools in Brisbane Australia. Children, and parents reported on children's social, emotional and behavioural characteristics at three-assessment points over 12 months. Results were examined universally (for all children) and for children who scored above the clinical cut-off for anxiety at pre-test. Results revealed that children in the FRIENDS intervention group reported fewer anxiety symptoms regardless of their risk status at posttest. In terms of reported levels of depression, only the high anxiety group who completed the FRIENDS intervention evidenced significant improvements at posttest. The results indicated that intervention gains were maintained over a period of 12 months, as measured by self-reports and diagnostic interviews. Moreover, evidence of a prevention effect was also demonstrated, with a significantly greater percentage of children in the control group progressed to "at risk" or "remained at risk" compared to children in the intervention group. Additionally, 85% of children in the intervention group who were scoring above the clinical cut-off for anxiety and depression were diagnosis free in the intervention condition at 12-month follow-up, compared to only 31.2% of children in the control group. Implications of these findings are examined, alongside limitations of the study and directions for future research.