Summary Objectives The aims of this study were to describe the epidemiology of melioidosis in tropical northern Australia and to assess the importance of defined risk factors. Methods The data were taken from a 14-year prospective study of 364 cases of melioidosis
in the ‘Top End’ of the Northern Territory. A whole-population logistic regression model was used to estimate the crude and adjusted relative risk (RR) for the defined risk factors. Results The mean age of the study population was 46.8 years, 264 (72.5%) were
male, 178 (49%) were aboriginal Australians and 59 (16.2%) died from melioidosis. Average annual incidence was 19.6 cases per 100 000 population, with an estimated rate of 260 cases per 100 000 diabetics per year. Using a whole-population logistic regression model, the estimated
crude and adjusted RR [95% confidence intervals (CI)] for melioidosis were 6.3 (5.1–7.8) and 4.0 (3.2–5.1) for those aged ≥ 45 years, 2.3 (1.8–2.9) and 2.4 (1.9–3.0) for males, 2.9 (2.3–3.5) and 3.0 (2.3–4.0) for aboriginal Australians,
21.2 (17.1–26.3) and 13.1 (9.4–18.1) for diabetics, 2.7 (2.2–3.4) and 2.1 (1.6–2.6) for those with excess alcohol consumption, 6.8 (5.4–8.6) and 4.3 (3.4–5.5) for chronic lung disease and 6.7 (4.7–9.6) and 3.2 (2.2–4.8) for chronic renal disease,
respectively. Conclusions Diabetes, excess alcohol intake, chronic renal disease and chronic lung disease are each independent risk factors for melioidosis. In tropical northern Australia, male sex, aboriginal ethnicity and age of ≥45 years are also independent predictors
for melioidosis. Impaired polymorph function may be critical in the predisposition to melioidosis.