The target population of the present study consisted of a 1-year (July 1985-June 1986) birth cohort from northern Finland. The prevalence of even slight hearing impairments (any threshold from 0.5 to 4 kHz > or = 25 db) at the age of 7 years was investigated among those 8,713 children still living in the area. The subjects for clinical audiometry were obtained in two ways. First, the standard clinical practice brought about 541 children, either with non-confirming results from their child welfare clinic screenings, suspected by their parents or already diagnosed as hearing impaired at a hospital. Secondly, in addition to this group, a random sample of 1,009 children, out of the 8,172 children not suspected, were also invited for audiometry. Of the clinical material, 101 children out of the 438 investigated were found to have impaired hearing according to the above criteria, and another 27 children out of the 789 investigated were obtained from the random sample. The estimated over-all prevalence of hearing impairments, even with the slight ones included, turned out to be 3.9% (95% confidence interval, CI, 2.7-5.7). Only 32% of the hearing impairments could be obtained according to the standard clinical policy! In conclusion, one cannot rely on clinical data when calculating prevalence figures for mild hearing impairments.