Completed West Africa is the poorest region of the world, and in most countries the majority of the population depends on rainfed agriculture. Chronic malnutrition is a constant feature of rural life in many areas, partly because agricultural productivity is so low. Technology is generally available to increase productivity, but a combination of political insecurity, failing infrastructure, and inadequate development of the institutions of civil society have made the transfer of knowledge to poor farmers slow or ineffective. For similar reasons, private sector service providers do not reach many remote, rural areas. In addition, labour migration has the consequence that women are the principal farmers in many areas, and are often discriminated against in other sectors such as health and education. Recent efforts to reform public sector research and extension services in the West African Region, and in Nigeria in particular, have been only partly successful. Participatory, problem-driven approaches often exist only in rhetoric in most government organisations (GOs). The pace of reform in GOs along the above lines is slow, and new approaches have considered the scope for partnership between RRIs (Regional Research Institutions), GOs, NGOs and, where possible and appropriate, farmers' organisations (FOs). It is claimed that Multi-Agency Partnerships (MAPs) such as these can reorient GOs by making demands on them, and complementing their technical skills with those of NGOs, in participatory needs assessment, screening of technical options, group formation and promotion of joint action. In the light of the increased emphasis on strengthening civil society, the climate is opportune for extending and expanding such partnerships. Knowledge systems must develop to reflect new conditions. A number of potential obstacles have to be negotiated if multi-agency partnerships are to serve as a successful link in the technology development dissemination and feedback chain. Power relations between NGOs and RRIs are unequal; they generally have differing philosophies, mandates, ethos, operating procedures, views on accountability, and external links with international organisations. In addition, the concept of monitoring process, which is essential for course-corrections if Multi-Agency Partnerships are to flourish, is widely unfamiliar to NGOs and NARS (national agricultural research systems). Five innovative techniques for incorporating national and international policy, and institutional issues into the design and delivery of NR research projects and programmes developed and promoted. Technical, administrative and managerial capacity of NGOs, NARs, RRIs and CBOs enhanced, in respect of forming and managing partnerships in pilot case study locations.Models of multi-agency partnerships developed, agreed by stakeholders and tested.Demand-driven, common-interest, local and regional networks strengthened through information-sharing.Approach disseminated widely within the region (through CORAF, NIRPANET and RADORT) and to the RNRKS (through the SEM Best Practice Guidelines series).