Screening for hearing defect by questionnaire and audiogram was undertaken in two group practices in the South-east of England in a total of 2,278 people aged 40-64 years.The prevalence of deafness in the combined populations of these two practices as measured by audiogram was found to be 5.8 per cent (+/- 0.5 per cent; one standard error). This figure is much more than estimates derived from American data for use in Britain. The prevalence of deafness was shown to increase with lower social class and age. The questionnaire was found to be unreliable for prevalence studies, but was thought to have some use as a screening instrument.There was an incidental finding of increased hearing defects in the left ears of men. The firing of rifles was felt to be a possible explanation of this.