Miombo woodlands supply ecosystem services to support livelihoods in southern Africa, however, rapid deforestation has necessitated greater knowledge of tree growth and off-take rates to understand the sustainability of miombo exploitation. We established 48 tree inventory plots within four villages in southern Malawi, interviewed representatives in these same villages about tree management practices and investigated the impact of climate on vegetation dynamics in the region using the ecosystem modelling framework LPJ-GUESS. Combining our data with the forest yield model MYRLIN revealed considerable variation in growth rates across different land uses; forested lands showed the highest growth rates (1639 [95% confidence interval 1594–1684] kg ha–1 year–1), followed by settlement areas (1453 [95% confidence interval 1376–1530] kg ha–1 year–1). Based on the modelled MYRLIN results, we found that 50% of the villages had insufficient growth rates to meet estimated off-take. Furthermore, the results from LPJ-GUESS indicated that sustainable off-take approaches zero in drought years. Local people have recognized the unsustainable use of natural resources and have begun planting activities in order to ensure that ecosystem services derived from miombo woodlands are available for future generations. Future models should incorporate the impacts of human disturbance and climatic variation on vegetation dynamics; such models should be used to support the development and implementation of sustainable forest management.