A total of 938 healthy 3-year-old children, divided into two unselected cohorts within the same geographic area, had tympanometric screening and re-screening. The results were assessed according to the parameters: season and the way the children were cared for. In all tests children in day-care centres had the highest prevalence of secretory otitis and tubal dysfunction due rather to an excess of long courses of illness than of an increased incidence. Regardless of the season, 8-13% of all children in day-care centres exhibited persistent secretory otitis through 6 months as compared with 2-4% of children not attending day-care centres. The lowest morbidity rate and the shortest course was found in children cared for at home whose tympanometric results were also to a marked extent independent of the season. Children in family day-care had results which in summer corresponded to those in children cared for at home and in winter approximately to those in the day-care centres. The theoretical and practical implications of the findings are discussed.