The ability of plant species to migrate is one of the critical issues in assessing accurately the future response of the terrestrial biosphere to climate change. This ability is confined by both natural and human-induced changes in land cover. In this paper we present land-cover and Carbon (C) cycle models designed to simulate the biospheric consequences of different types of land-cover changes. These models, imbedded in the larger integrated assessment model IMAGE 2, were used to demonstrate the importance of considering spatial aspects for global C-cycle modelling. A gradual-migration, an unlimited-migration and a no-migration case were compared to show the range of possible consequences. Major differences between these cases were simulated for land-cover patterns and the carbon budget. A large geographical variation in the biospheric response was also simulated. The strongest response was simulated in high-latitude regions, especially for the migration cases in which land-cover changes were permitted. In low-latitudes regions the differences between the migration cases were smaller, mainly due to the effects of land-use changes. The geographical variation among, and the different responses, the migration cases clearly demonstrate how essential it is to assess biospheric responses to climate change and land use simultaneously. Moreover, it also shows the urgent need for enhanced understanding of spatial and temporal dynamics of the biospheric responses.