The Advanced Very High Resolution Radiometer (AVHRR) is currently used for vegetation monitoring. Information on the state and evolution of the vegetation cover on continental areas is routinely obtained from Global Vegetation Index data (GVI). In this paper some aspects of continental level, biome features were interpreted from a global scale GVI time series acquired with the NOAA/AVHRR between 1986 and 1990. A biome classification scheme for Africa is proposed. NOAA/AVHRR data were submitted to a cloud filtering algorithm, subsequently calibrated and corrected for atmospheric perturbations. Typical behaviour of Normalised Difference Vegetation Index (NDVI) temporal evolution derived from the AVHRR are classified, colour coded, and mapped to be compared with classical biome cartographic information. Analysis of the temporal GVI shows, that the spatial organisation of natural biomes in Africa changes within a single growing season. Transition zones between major ecological zones can clearly be identified. Furthermore, closer examination of GVI temporal variations, reveals year to year differences assumed to be the result of changes in landscape structure. Results illustrate that this approach is appropriate to describe vegetation temporal dynamics. The approach shows potential to describe climatic and human interactions with African biomes. In this paper results are presented and limitations will be discussed.