The paper discusses the meaning and scope of agroforestry, and reasons that subsistence of rearing of unevenly scattered and spares natural trees, the sole cropped from forestry, and the Taungya system can not be considered agroforestry. Agroforestry systems, on the contrary, may include - parts of farms comprising rocky, steep, and poor soil areas used for raising fuelwood, or small timber/pole, or fodder trees while better areas are used for seasonal crops; - parts of farms sequentially used for tree and for seasonal crops to enhance total outputs and variety of goods, and to minimize climate related risks; - farm corners/bunds/irrigation channels used for tree crops to meet family needs for fuelwood, fodder, small timber, etc - Agrisilvi systems in situations with strong pastoral base but restrictions on open grazing The intuitive reasoning behind usually perceived benefits of agroforestry: enhanced produtivity of the land resource, enhanced employment opportunities for the rural labourers, reduction in seasonal vulnerability for supplies of essential economic goods, enhancement in stock of productive assets in the form of trees, etc have been summarized. The paper then discusses the methodology for and the data required for systematic assessments of the impacts of agroforestry systems. Changes in annuities, and changes in wage rates along with changes in total employment and seasonal employment patterns have considered as composite and relatively simple indicators. Assessments of potential economic impacts at the regional and national levels could be made through income-investment multipliers, but the data requirements would be extremely difficult to meet. Alternatively, indicators of agroforestry could be developed both in financial and benefit-cost terms, and the latter could be used for policy purposes. The paper also highlights some general issues for consideration of the policy makers and researchers.