Global change impacts on biodiversity, ecosystem processes and human well-being are a pressing scientific and political concern. One of the major goals of global change research is to reliably predict the future distributions of species and to assess potential resulting consequences for the adaptive management of ecosystems and their services and for conservation strategies. Dramatically increased availability of biodiversity data at broad scales, recent advances in statistical approaches such as ‘species niche modelling’ and progress in high-performance computing now allow data handling and the simulation of future climatic and biodiversity scenarios. Despite the popularity of niche modelling for predicting the potential impacts of global change on species distributions, to date a validation of this approach at biogeographically relevant scales is lacking. Through the use of very extensive datasets of North American birds (starting in 1966), this project will assess the reliability of niche modelling predictions through cross-temporal validations. Such independent validation will also enable us to assess which predictors and what spatial scale provide most accurate predictions. Using this knowledge, I will then predict the potential impact of global change on North American bird distributions. Going beyond just species richness, I will also evaluate and predict changes to the functional and phylogenetic make-up of bird communities. Finally, where possible I will carefully extend several of these predictions of global change impacts to breeding and wintering communities of birds worldwide. This first global assessment will e.g. help identify, at biogeographic scales, migratory birds species potentially most affected by impending climatic changes or pinpoint areas where the bird diversity may be particularly exposed to future change. I expect the resulting findings to significantly advance broad-scale global change ecology and help inform conservation and policy.