Chronic otitis media (OM) is an archetypal complex disease, which is particularly prevalent in childhood. Epidemiological data suggest high heritability for disease susceptibility, but previous genetic association studies have had methodological flaws, and none have specifically focused on chronic OM phenotypes. Mouse models represent one way to ascertain candidate loci for human association testing. A number of mouse models of middle ear inflammation have been reported, but many susceptibility loci remain undiscovered. I demonstrate that oto-endoscopy is a robust and scalable phenotyping platform for OM in the mouse, and discuss its value in new model discovery. Chronic OM is also a feature of trisomy HSA21 (Down Syndrome). Through an interrogation of the mouse library of segmental trisomy models of Down Syndrome, I identify a critical trisomic region for chronic otitis media. This region may underlie OM susceptibility in Down Syndrome, but could also contribute to disease susceptibility in non-syndromic disease. Mouse models can also be used to interrogate disease mechanisms. Our previous work has shown that the chronically inflamed middle ear is hypoxic, and that hypoxia signalling is a potential therapeutic target. Exploiting the Junbo mouse model, I demonstrate that surgical ventilation of the Junbo ear improves inflammation, and that this is associated with loss of hypoxia signalling. I present preliminary results from transcript analyses of human middle ear effusions showing marked upregulation of hypoxia signalling. A systematic review of existing mouse models suggests that the loci FBXO11, EVI1, SMAD2, and TGIF1 are good candidates genes for human association testing. I detail recruitment and collection of DNA from families in the UK where a child is undergoing grommet insertion. Association testing using a variant of the transmission disequilibrium test shows susceptibility associated with polymorphisms at FBXO11, and possibly also SMAD2 and TGIF1.