A brief synthesis of a longer report. Provides a micro-level foundation for discussions of land allocation and its relation to poverty within the smallholder sector of Eastern and Southern Africa based on results from household surveys in Ethiopia, Kenya, Rwanda, Mozambique and Rwanda between 1990 and 2000. Addresses (1) why geographically based targeted approaches to poverty reduction are likely to miss a significant share of the poor, (2) why agricultural growth alone is not likely to be a sufficient engine for directly lifting a significant share of small-scale farmers out of poverty, (3) why agri-food productivity growth is needed to create a more dynamic and diversified rural economy, (4) why increased access to land is likely to affect significantly the poverty-reducing effects of agricultural growth. Concludes with implications for the design of poverty reduction strategies.