OBJECTIVE: To determine the long-term hearing following surgical treatment of chronic OME in early childhood (myringotomy or ventilation tube) and to determine the impact of the occurrence and the extension of specific eardrum pathology on the hearing level. PATIENTS AND METHODS: In 1977-1978, 224 consecutive children (91 girls and 133 boys) with chronic bilateral OME were enrolled and treated by adenoidectomy, bilateral myringotomy and insertion of a ventilation tube on the right side only. In 2002, a follow-up examination included otomicroscopy and audiometry. Hearing thresholds were compared to an age- and gender-matched normative data set. For the determination of the impact of specific eardrum pathology on the hearing, multiple linear regression modelling was used in adjustment for age and concomitant eardrum pathologies. RESULTS: Long-term hearing after chronic OME and associated treatment is not different from age and gender-matched normal hearing. Treatment modality (myringotomy or ventilation tube) has no impact on the long-term hearing level. The regression analyses showed that the presence of myringosclerosis is associated with an overall hearing loss in myringotomised ears (4-5dB), but not in tubed ears, for which only high frequencies were affected. Conversely, tensa atrophy is associated with an overall hearing loss in tubed ears (3-4dB), but not in myringotomised ears, for which only high frequencies were affected. CONCLUSIONS: Hearing 25 years after surgical treatment of chronic OME is not different from age and gender matched normal hearing. In addition, treatment modality (myringotomy or ventilation tube) has no impact on the long-term hearing level. However, atrophy is associated with a hearing loss in tubed ears, whereas myringosclerosis is associated with a hearing loss in myringotomised ears. The hearing loss associated with pathology is of limited magnitude (up to about 5dB PTA). Even though pathology does occur more frequently and more extensively in tubed ears, the effect on the hearing at the group level is too small to have an impact. It is important to note, that this may be due to a type 2-error (number of patients too small to show an effect).