Agricultural expansion and over-cutting of trees for fuelwood are important causes of deforestation in arid and semi-arid countries such as Sudan. The consequence is increased desertification and high erosion and loss of soil nutrients leading to declining agricultural productivity. However, the social costs of the deforestation externality are not taken into account in present forest management and land use planning in Sudan leading to under-pricing and over-exploitation of the country's forest resources. This study evaluated the suitability of approaches commonly used by most forest resource management agencies for prediction of the state and control of harvesting of forest resources against alternative empirical simulation models using relevant information about economic behaviour of trading agents in the fuelwood market. Results showed the clear superiority of models integrating market behaviour over current approaches in the ability to better simulate real trends of wood consumption and hence depletion rates. The study also adopted an optimal control model to derive socially optimal forest harvesting regimes. The results showed that current rates of forest resource rent recovery and reforestation efforts are very far from optimal. Results also suggest that, in addition to optimal pricing and higher reforestation efforts, promotion and availability of fuel substitutes and investment in wood energy conversion efficiencies have a strong potential for curbing the problem of deforestation in Sudan.