ObjectiveIn children with otitis media with effusion (OME), to investigate the incidence of, and any association, between retractions of the pars tensa and pars flaccida; to assess the effect of pars tensa and pars flaccida retractions on the hearing; to investigate risk factors for retractions; and to document the natural history of such retractions over a 12-week “watchful waiting” period.Study DesignProspective, observational study.SettingSixteen departments of otolaryngology in hospitals in the U.K.PatientsA cohort of 1,267 children aged 3.25 to 6.75 years with confirmed OME. None had previously received surgical intervention.InterventionsFollow-up over a “watchful waiting” period of 12 weeks.Main Outcome MeasuresOtoscopy and pure-tone audiometry.ResultsRetraction of the pars tensa to the incus or promontory occurred in 8% of the better-hearing ears and 10% of the poorer-hearing ears. Pars flaccida retraction to the malleus or farther occurred in 4.5% of the better-hearing ears and 5.5% of the poorer-hearing ears. Retractions were not associated with a longer history of hearing problems. Pars tensa or pars flaccida retraction in association with OME did not materially affect the hearing. Pars tensa retractions, followed up over a 12-week period, resolved in 69% of the better-hearing ears and 65% of the poorer-hearing ears. In 14% and 10% of ears, respectively, the OME had also resolved.ConclusionsThere is minimal evidence to support the concept that pars tensa or pars flaccida retractions are a strong or relevant marker for the severity or evolution of OME in children. Prospective studies over a longer period of follow-up are required to confirm this.