Important policy steps have been taken in the past 25 years to strengthen the national agricultural research systems (NARS) of many developing countries. The paper discusses the organization and structure of these systems in different parts of the world. They include the newly created agricultural research councils of Asia, the national institutes of Latin America, and the ministry and university-based research services. The emerging NARS of sub-Saharan Africa and those found in the Arab world have been reviewed in this context. The discussion has been organized in relation to the potentials which have been created for the improved performance of the research and governance functions of these systems. Many of the reorganized systems have already made considerable impact as a result of their newly created potential. Many others are in early stages of their evolution and would need further strengthening before making a impact of this kind. It has been argued that the process of reorganization must continue for all of them so that they become better equipped for the new challenges which will arise from continued population pressures and the needs of economic development. The nature of this new reorganization has been considered. It involves, among other things, creation of greater capacity for planning so that the NARS leaders move away from routine administration to policymaking. Some of the other important issues discussed in this context are the need for stronger links between development and research and the structural changes which will be needed to forge these links; decentralization and regional research; and inter-institutional coordination at the national level. It has been stressed that the heavy investments made by the governments in the 1960″s and 1970″s may not continue and the demand in future will be for greater returns from investments already made through better planning and greater efficiency in the management of resources. The organization of research station networks has been considered in the wider contex of the organization of NARS. The rationale for the different types of stations and their number has been examined. Finally, the internal structure of the research stations has been discussed as a function of their mandate.
