BACKGROUND AND AIMS: A robust climate-health projection model has the potential to improve health care resource allocation. We aim to explore the relationship between Australian intensive care unit (ICU) demand and various measures of the long-lived large-scale climate and to develop a future nationwide climate-health projection model. METHODS: We investigated patients admitted to ICUs in Australia between January 2003 and December 2019 who were exposed to long-lived large-scale combined climatic measures of temperature and humidity. We analysed the projected demand for respiratory-related ICU average length of stay (in days) per capita (ICU(D/C)) with four historical and one future projection dataset. These datasets included: i) Australian and New Zealand Intensive Care Society adult patient database, ii) Socioeconomic Data and Applications Center gridded global historical population, iii) Australian Bureau of Statistics national historical population, iv) Japanese 55-year Reanalysis historical climate (JRA55), and v) the fifth Coupled Model Inter-comparison Project future climate projections. RESULTS: 148,638 patients with respiratory issues required intensive care between 2003 and 2019. The annual growth in the population density-weighted wet-bulb-globe temperature-a combined measure of temperature and humidity-is strongly correlated with the annual per capita growth ICU(D/C) for respiratory-related conditions (r=0.771; p<0.001). This relationship was applied to develop a model projecting future respiratory-related ICU demand with three possible future Representative Concentration Pathways (RCP). RCP2.6 (lowest carbon emission climate scenario) showed only a 33.4% increase in Australian ICU(D/C) demand by 2090, while the RCP8.5 (highest carbon emission climate scenario) demonstrated almost two-fold higher demand (66.1%) than RCP2.6 by 2090. CONCLUSIONS: The annual growth in population density-weighted wet-bulb-globe temperature correlates with the annual growth in Australian ICU(D/C) for respiratory-related conditions. A model based on possible future climate scenarios can be developed to predict changes in ICU demand in response to CO(2) changes over the coming decades.