Background: Climate change will probably alter the spread and transmission intensity of malaria in Africa. Objectives: In this study, potential changes in the malaria transmission are assessed via an integrated weather-disease model. Methods: We simulated mosquito biting rates by the Liverpool Malaria Model (LMM). The input data for the LMM were bias-corrected temperature and precipitation data from the Regional Model (REMO) on a 0.5 degrees latitude-longitude grid. A Plasmodium falciparum infection model expands the LMM simulations incorporating information on the infection rate in children. Malaria projections were carried out with this integrated weather-disease model for 2001-2050 according to two climate scenarios that include the effect of anthropogenic land use and land cover changes on climate. Results: Model-based estimates for the present climate (1960-2000) are consistent with observed data for the spread of malaria in Africa. In the model domain, the regions of epidemic malaria occurrence are located in the Sahel as well in various highland territories. A decreased spread of malaria over most parts of tropical Africa is projected due to simulated increased surface temperatures and a significant reduction in annual rainfall. However, the likelihood of malaria epidemics is projected to increase in the southern part of the Sahel. In most of East Africa, malaria transmission intensity is expected to increase. Projections indicate that highland areas that were formerly unsuitable for malaria will become epidemic, while in the lower altitude regions of the East African highlands, epidemic risk will decrease. Conclusions: We project that greenhouse gas and land use driven climate changes will significantly affect the spread of malaria in tropical Africa well before 2050. The geographic distribution of epidemic malaria areas might be strongly altered in the coming decades.