The adaptation of agriculture and forestry to the climate of the twenty-first century supposes that research projects will be conducted cooperatively between meteorologists, agronomists, soil scientists, hydrologists, and modellers. To prepare for it, it is appropriate first of all to study the variations in the climate of the past using extensive, homogenised series of meteorological or phenological data. General circulation models constitute the basic tool in order to predict future changes in climate. They will be improved, and the regionalisation techniques used for downscaling climate predictions will also be made more efficient. Crop simulation models using input data from the general circulation models applied at the regional level ought to be the favoured tools to allow the extrapolation of the major trends on yield, consumption of water, fertilisers, pesticides, the environment and rural development. For this, they have to be validated according to the available agronomical data, particularly the available phenological series on cultivated crops. In addition, climate change would have impact on crop diseases and parasites, as well as on weeds. Very few studies have been carried out in this field. It is also necessary to quantify in a more accurate way the stocks and fluxes of carbon in large forest ecosystems, simulate their future, and assess the vulnerability of the various forest species to a change in climate. This is all the more important in that some propagate species choices must be made in the course of the next ten years in plantations which will experience changed climate. More broadly speaking, we shall have not only to try hard to research new agricultural and forestry practices which will reduce greenhouse gas emissions or promote the storage of carbon, but it will also be indispensable to prepare the adaptation of numerous rural communities for the climate change (with special reference to least developed countries in tropical areas, where ma