A rapid increase in the productivity of all sectors is required in developing countries for the betterment of their people and removal of abysmal poverty. This may be achieved through the institution of planning at the national, sectoral and regional levels. Proper choice of projects to be included in the plan, their location and regional distribution play an important role in increasing the productivity of planned sectors of the economy and in the development of various regions. This paper discusses the application of operations research models in the dual hierarchial framework by regions and sectors of a national economy. A class of computational techniques known as decomposition methods exhibit hierarchical problem structure quite similar to the multi-level planning problem. Detailed sectoral models or regional models involving project selection and spatial details could be integrated into a balanced national plan using such a planning and computational framework. An analysis of the various commodities produced would indicate the level at which project decisions for that commodity should be undertaken. The partial decomposition of more amenable to the application of operations research models for project selection and optimization.