The dispersion of radionuclides in dust and inhalation dose rates to the public from the planned remediation of the Ranger uranium mine in the wet-dry tropics of Australia was modelled using RESRAD-OFFSITE. Dust inhalation dose rates were predicted to be highest on the remediated site and decrease with an approximate inverse square to inverse cubic dependence with distance from the site. The annual dose above natural background to a hypothetical individual permanently occupying the remediated site (representing the worst case scenario for radionuclide in dust exposure) was estimated to be 5.3 × 10-3 mSv. The estimated doses from exposure to radionuclides in dust were two to three orders of magnitude lower than those from exposure to 222Rn. A sensitivity analysis showed that source-related and receptor-related model parameters had direct proportional influences on dust inhalation dose rates. Four transport-related model parameters (atmospheric stability class, deposition velocity of particulates, precipitation and wind speed) were also influential and generally had an increasing influence with distance from the source. The results of this study may provide general guidance to similar sites elsewhere on the relative importance of dust versus gaseous 222Rn transport pathways and the relative influence of dispersion modelling parameters on predicted exposures and doses.