The problem of effectively managing patients with functional somatic symptoms remains huge in both primary and hospital care, although the potential bene ts of mental health interventions such as counselling or psychotherapy are still not widely accepted. Unfortunately, the cost-effective evaluation of counselling has been beset by methodological problems, which stem from attempting to make comparisons between counsellors with different types of training, therapeutic approach and case mix. Since these real-world variations in counselling practice are likely to remain, it is proposed that a different approach to the evaluation of cost-effectiveness should be adopted in the hope of accelerating acceptance of the value of psychological interventions for the army of patients who are currently labelled as 'dif cult' or 'heartsink' and who do not receive the treatment they deserve from the health-care services. It is proposed that, since the common aim of almost all mental health interventions is to engender or enhance a sense of 'personal responsibility' in patients/clients, this core objective should form an agreed 'yardstick' against which to evaluate the impact of various mental health interventions. Furthermore, since personal responsibility and future health-care utilization (and associated costs) can be assumed to be directly causally related, it is proposed that offset costs should form the basis for future cost-effectiveness evaluations.