OBJECTIVES : To examine the long-term effects of predominant breastfeeding on incidence of otitis media. DESIGN :Prospective birth cohort study. SETTING : The West Australian Pregnancy Cohort (Raine) Study recruited 2900 mothers through antenatal clinics at the major tertiary obstetric hospital in Perth, Western Australia between 1989 and 1992. PARTICIPANTS : 2237 children participated in a six year cohort follow-up and a subset of 1344 were given ear and hearing assessments. MAIN OUTCOME MEASURES : OM diagnosis at six years of age (diagnosed by low-compliance tympanograms, 0 - 0.1 mmho). This was compared to OM diagnosed at the three year cohort follow-up using parent-report measures. Main exposure measures were duration of predominant breastfeeding (defined as the age other milk was introduced) and duration of partial (any) breastfeeding (defined as the age breastfeeding was stopped). RESULTS : There was a significant, independent association between predominant breastfeeding (OR = 1.33 [1.04, 1.69]; p = 0.02) and OM, and breastfeeding duration (OR = 1.35 [1.08, 1.68]; p = 0.01) with OM at three years of age. However, at six years of age this relationship was no longer statistically significant (predominant breastfeeding OR = 0.78 [0.48, 1.06]; p = 0.09; duration of breastfeeding, OR = 1.34 [0.81, 2.23]; p = 0.25). CONCLUSIONS : Our findings are in line with a number of epidemiological studies which show a positive association between breastfeeding and OM in early childhood. However, the long-term follow-up of these children revealed that by six years of age there was no significant influence of breastfeeding on presence of OM. These results suggest that the protective effect of predominant breastfeeding for at least six months does not extend to school-age children, where other social and environmental factors may be stronger predictors of OM.