We evaluated an Internet-based psychological intervention supported by either general practitioners or psychologists (Panic Online), and a Primary-care Evidence-based Psychological-interventions (PEP) strategy which involves training GPs to deliver specific psychological interventions. Economic modelling suggests that Panic Online is cost-effective when supported by either GPs or psychologists. Threshold analysis of the psychological training of GPs suggests that a modest effect size for clinical benefit would be sufficient to provide an acceptable cost-effectiveness ratio. The sustainability of these approaches depends on a range of factors, including funding, workforce availability, and acceptability to consumers and health care providers.