Objectives: To measure the impact of 4 weeks of daily swimming on rates of ear discharge among Aboriginal children with a tympanic membrane perforation (TMP) and on the microbiology of the nasopharynx and middle ear. Design, setting and participants: A randomised controlled trial involving 89 Aboriginal children (aged 5-12 years) with a TMP, conducted in two remote Northern Territory Aboriginal communities from August to December 2009. Intervention: 4 school weeks of daily swimming lessons (45 minutes) in a chlorinated pool. Main outcome measures: Proportions of children with ear discharge and respiratory and opportunistic bacteria in the nasopharynx and middle ear. Results: Of 89 children randomly assigned to the swimming or non-swimmin, groups, 58 (26/41 swimmers and 32/48 non-swimmers) had ear discharge at baseline. After 4 weeks, 24 of 41 swimmers had ear discharge compared with 3 of 48 non-swimmers (risk difference, -8% (95% CI,-28% to 12%). There wei no statistically significant changes in the microbiology of the nasopharynx or middle ear in swimmers or non-swimmers. Streptococcus pneumoniae and nor typeable Haemophilus influenzae were the dominant organisms cultured from the nasopharynx, and H. influenzae, Staphylococcus aureus and Pseudomonas aeruginosa were the dominant organisms in the middle ear. Conclusions: Swimming lessons for Aboriginal children in remote communitie should be supported, but it is unlikely that they will substantially reduce rates of chronic suppurative otitis media and associated bacteria in the nasopharynx and middle ear. However, swimming was not associated with increased risk of ear discharge and we found no reason to discourage it.