Dire consequences for global freshwater biodiversity have been predicted in view of expected climate change over the coming century. The Australian freshwater biota, exposed to periods of severe aridity in the past, may have evolved wider resistance and resilience to hydrological fluctuations than that in many parts of the world. However, it also includes many cold-adapted species that are highly susceptible to increasing temperatures. Forecasting of species’ responses to projected climate change has been heavily dominated by correlative bioclimatic modelling of geographic distributions, but Australian freshwater species may undergo a variety of climate-related impacts and adaptations other than range shifts. Population responses during the recent Millennium Drought and the subsequent wetter and cooler phase have differed greatly among freshwater taxa, exemplifying the likely complexity of reactions to future climates. This variability can be explained to some degree by existing knowledge of species traits and life history strategies, indicating the potential for mechanistic predictions of how species will respond to future climatic trends and extremes. I suggest that correlative forecasting needs to be complemented by a greater understanding of the biology and ecology of sensitive species, including interspecific interactions, leading to more process-based forecasting. Mechanistic modelling should also help to reduce uncertainty about the benefits and risks of actions that are frequently proposed for conserving freshwater biodiversity in the face of climate change, such as protecting refuges, increasing connectivity, use of environmental water allocations, stream shading and managed relocation.