Background: Melioidosis correlates strongly with rainfall, but there is substantial di-versity in climate between melioidosis-endemic locations. The Northern Territory of Aus-tralia epitomises the “wet/dry” tropics, with a prolonged dry season from May to October. We analysed dry season cases of melioidosis during 35 consecutive years and compared these with wet season cases. We aimed to provide insights into how dry season cases of melioidosis may occur in this region. Methods: Case epidemiological and clinical data were extracted from the Darwin Prospective Melioidosis Study. Weather parameters, in-cluding daily rainfall, were analysed using generalised additive models and conditional logistic regressions to assess associations between dry season cases and preceding rain-fall. Results: Of 1520 melioidosis cases between 1989 and 2024, there were 325 (21%) in the dry season. While the well-recognised clinical diversity of melioidosis was also seen amongst dry season cases, pneumonia was proportionally less common and cutaneous melioidosis was more common than in the wet season. A higher proportion of dry season patients (23%) had no identified clinical risk factors for melioidosis, and mortality in the dry season was 8%. There was a range of plausible explanations for many of the dry sea-son cases, including unseasonal rainfall prior to infection. Infections in urban settings were notable, with anthropogenic factors such as irrigation and construction resulting in persistence of Burkholderia pseudomallei in the environment during the dry season. 3% of cases remained unexplained. Conclusion: Further prospective studies are needed to better define the infecting events resulting in melioidosis, especially in the dry season. These studies should include a timely history taking from the case and their family and selected environmental sampling for B. pseudomallei.