Carbon sequestration and avoided CO2 emissions from deforestation are hotly debated issues in the negotiations under the United Nations Framework Convention on Climate Change. In this paper a new version of the global forest model (G4M) is described and used for determining optimal carbon prices for slowing down deforestation and stimulating afforestation. The global land-use model GLOBIOM predicts price changes for land and forest products, which are used in G4M to determine afforestation and deforestation patterns in geographic space. We find global average cost efficient carbon price to be 25$-tC-1 causing sequestration of 15 Gt of carbon per 50 years in aboveground phytomass additionally. Top five countries by potential of sequestration of additional carbon and cost competitiveness are Brazil, Zaire, Indonesia, Bolivia and Tanzania. It is shown that changes in agricultural and forestry markets influence the competition for land and may modify the effect of economic incentives for carbon sequestration considerably. The model results coincide well with independent projections made for Ukraine.