Some authors argue that deforestation in Africa is one of the highest among the world's continents. However, for several regions, historical data and observations obtained by remote sensing raise some controversy over the dramatic portrait accepted in international circles. The case study, located in the center plateau of Angola, highlights the importance of multi-temporal observations of land cover by remote sensing as a means of obtaining objective information about deforestation. This work analyses land cover change in the province of Huambo, between 1990 and 2000, as a basis for developing a monitoring system that supports sustainable forest management. For that, we used Landsat TM images, and produced land cover maps using a maximum likelihood algorithm. Results show a small increase of Closed-Miombo and Open-Miombo and a decrease of Savanna-Woodland, while Agriculture expands slightly. These results are discussed jointly with the socio-economic factors that may have driven them.