The major epidemiological investigations of hearing impairment, disability and handicap show that the elderly are the group most disabled by their hearing impairment. There is considerable debate concerning the most efficient way of reducing this inevitable burden of age-related hearing impairments in the next generation. Early fitting of 'targeted' individuals with hearing aids may help but there are a large number of methodological problems associated with conducting and evaluating such a programme of research (especially retrospectively). The logical prerequisite to early fitting as a means of reducing later disability is to ascertain the acceptability of and benefit given by intervention at this early stage. This study therefore set out to investigate the age/sex register provided by the primary physician (GP) as an appropriate base to identify candidates for early aid fitting among a sample of middle-aged patients (50-65 years) living in Roath, Cardiff. Of the 662 who replied to an initial contact letter (1050 were on the age/sex register), 21 already possessed hearing aids. After screening and examination 66 people were offered some form of management which was accepted by 43 during the course of the study. Aid use thereby increased from about 3% to over 9% in this middle-aged group. A 2 year follow-up indicated continued use of the aids, and benefit on a speech reception task was measured. The cost of detecting those who might benefit was calculated using a two-question 'paper and pencil' screen as the first step. A national programme for Wales would cost at least 188,000 pounds per annum at 1990 prices over an initial 5 year span if a criterion which aimed to find at least 45 dB HTL impairments over mid-frequencies was implemented. For a criterion of 35 dB the cost would be 378,000 pounds pa.