Researches on the projection of climate change are comprehensively reviewed here. Firstly, three kinds of climate models are presented in order to evaluate the ability of models (in particular coupled atmosphere-ocean general circulation models) in the application of climate change projection. It is outlined that coupled models can provide credible simulations of the present climate for most variables of interest for climate change although clouds and humidity remain significant sources of uncertainty. Secondly, a series of emission scenarios released by the IPCC are summarized, so are the results of climate change projections at the global scale and in East Asia. Analysis indicates that in spite of the different outputs from global climate models (GCMs) with various scenarios, the model experiments show similar climate pattern with global warming and the regional increasing of mean precipitation over the 21st century. Considering the coarse spatial resolution of GCMs for many subgrid scales, this paper introduces the downscaling techniques used to enhance the regional details of climate projection. Then, the uncertainties of climate change projection are addressed. At last, the perspectives of research on the future climate change projection as well as the priority focus of climate change projection in China are discussed.