A genetic contribution to infectious disease in human populations has long been suspected and is now supported by more than 50 years of epidemiological evidence showing, for example, infection rates to be much higher than disease rates. In successful family studies of high-penetrance effects, single gene mutations have been identified that reveal a molecular mechanism leading to increased risk of a specific infectious disease. However, in population-based studies, genetic variants conferring host susceptibility to various infectious diseases have been difficult to uncover. Although mutations such as that in the CCR5 gene, which confers protection against HIV infection, have been reliably discovered, polymorphisms affecting larger proportions of a population have been hard to prove definitively. The recent arrival of the genome-wide association study format, currently being applied to Kawasaki disease, tuberculosis, malaria, HIV, dengue and others, gives us hope that these challenges can finally be met, with implications for population-based treatment and prognosis strategies.