Sri Lanka Agricultural Research Project, supported by Credit 1776 for US$18.6 million, was approved in FY86. The credit was closed on December 31, 1996, as planned. A total of US$3.7 million was canceled in November 1995, due to under-spending on contract research and incremental staff costs, and an undisbursed sum of US$0.6 million was canceled on May 15, 1997. Cofinancing of US$3.2 million was provided by Germany’s GTZ. The Implementation Completion Report (ICR) was prepared by the South Asia Regional Office. The borrower’s and cofinancier’s contributions are included as appendices. The project objective was to raise farmers’ incomes by increasing agricultural production through the development of improved crop varieties and improved husbandry practices and farming systems. The project had eight components: (i) establishment of a Council for Agricultural Research Policy (CARP) with authority to formulate a National Agricultural Research Plan (NARP); (ii) provision of infrastructure and equipment to research institutions; (iii) staff development and technical assistance; (iv) the introduction of a contract research program; (v) establishment of a research planning cell and project management unit; (vi) provision for incremental staff and operating costs; (vii) introduction of research planning and evaluation procedures; and (viii) improvement of research-extension linkages. The project was partially successful in meeting its objectives. The physical targets were substantially met, but there was less success with institutional development. CARP was established with statutory authority to formulate the NARP. However, a number of institutional and functional issues prevent CARP from being able to seriously influence the focus or the resource allocation of the National Agricultural Research System. By 1996 it was able to produce only a draft NARP, an important early priority. Virtually all civil works have been completed, although with delays, to satisfactory design and construction standards. The staff training program exceeded its targets for M.Sc./M.Phil. fellowships and short-term training, creating a cadre of capable researchers with significant potential payoffs in the long run. The technical assistance was directed mainly at assisting in the establishment of CARP and strengthening the research institutes. It also helped support the development of management tools for research priority setting and planning, and the production of NARP. The contract research program also faced considerable delays, with problems including poor research prioritization, the mixed quality of research, and, until recently, the lack of a monitoring and evaluation system. The project funded incremental staff at research institutes and stations. All positions were filled, albeit with delays, but with recurring problems in filling positions for agricultural economists. Research-extension linkages remain poor. Socioeconomic studies were not conducted for lack of agricultural economists and the adaptive research program was only partially implemented due to limited training and use of contract research funds. In agreement with the ICR, OED rates outcome as marginally satisfactory, institutional development as modest, sustainability as uncertain, and Bank performance as satisfactory. The ICR draws some familiar lessons of which the most important are: economic analysis should be introduced for research projects; national research systems should institute performance indicators and evaluation systems that track resource allocation, monitor progress against research objectives, and identify impact on national policy objectives; to strengthen research-extension linkages, adequate staffing, training, and procedures for needs identification, beneficiary participation and monitoring and evaluation are required. The ICR is satisfactory. No audit is planned.