The role of otitis media (OM) was investigated in a group of preschool children with a rigorously defined language impairment: developmental expressive language disorder (ELD). No differences in the frequency, duration, or timingof medically documented episodes of OM were found between 65 ELD and 50 normal children. However, longitudinal analyses with the ELD children revealed a relation between OM and expressive language improvement. The strongest positive predictor of language improvement between 2 and 3 years of age was the amount of time spent with OM between 12 and 18 months of age. These findings suggest that children with ELD represent 2 etiological populations depending on a history of OM. When ELD is influenced strongly by OM during a critical period for the development of expressive language, improvement is highly likely. However, in the absence of such a history, the causes of ELD are likely to be more central and long-lasting. The absence of between-groups differences, coupled with the predictive power of OM history within the ELD group, suggests either that another factor interacts with OM to cause ELD or that OM depresses development for all children.Otitis media (OM), or inflammation of the middle ear, is a common disease of infancy and early childhood. Approximately 50% of all children experience at least one episode of OM during the 1st year of life, and by 3 years of age, approximately two thirds of all children experience at least one episode; one third of all children experience three or more episodes of OM prior to age 3 (Howie, 1980;Teele, Klein, & Rosner, 1980. During the acute phase of OM, children may experience pain, fever, and other overtly expressed symptoms. OM is typically associated with fluid in the ear (effusion) that persists after the acute phase of the infection has passed (Bluestone, 1984;Marchant et al, 1984;Teele et al., 1980Teele et al., ,1984. The occurrence of OM is often associated with conductive hearing loss of a transient nature and may lead to chronic auditory impairment (Bess, 1983;Fria, Cantekin, & Eicher, 1985; Wright etal., 1988).Several studies have found a relation between an early history of OM and later problems in language development. For instance, Teele et al. (1984) followed a cohort of 205 children from birth to 3 years of age. The presence of OM was periodically assessed by otoscopic examination. Tests of speech and