The continued appearance of new screening tests for dementia suggests both the continuing need for a valid measure of pathological cognitive change and dissatisfaction with what is currently available. A brief review of current tests is presented which focuses on the major theoretical and methodological issues which have impeded the development of adequate screening instruments. Strategies for the management of such difficulties are also indicated, with a view to guiding prospective researchers in this field and increasing the critical perspective of those currently using these instruments.