Summarizes attached interim evaluation of a project to improve farmer utilization of water and agricultural recovery of fertile valley lands in the southern regions of Ziguinchor and Kolda, Senegal. The evaluation covered the period 1988-2/94. After a slow start, the project has begun to achieve many of its purposes and objectives, though lack of baseline data precludes reliable assessment of the project's impact with regard to areas of land reclaimed and increase in rice production. Further, the project is not achieving its ambitious level of outputs. The project cannot achieve its goal and purpose by 1996 with water resources management alone; unless an agricultural extension and credit component is added, the project will not become economically viable. Due to engineering requirements, 3-meter crest dikes were constructed in lieu of 1-meter crest dikes. Implementation has also been constrained by the failure of several project paper assumptions to materialize: a secure political situation; favorable conditions for cereal marketing; the existence of financial incentives for increasing cereal production; the availability of extension services and of skilled civil servants (especially topographers); easy access to and use of improved inputs; lack of labor constraints; timely procurement of construction materials; and the rapid build up of local private sector and farmer expertise (in particular the ability to build dikes). Further, villagers seem to lack a clear idea of the roles, duties, and functions of the Village and Inter-Village Water Management Committees, and there is no trained rural sociologist to oversee village organizations and to supervise the project sociology team and NGO personnel working in village organizations. The following lessons were learned. (1) The project goal was overestimated because of a false assumption that water control management alone will promote the interest of farmers, who will then automatically use agricultural inputs to increase rice production. (2) Careful study should be made in costing infrastructure and in assessing the availability of necessary construction capabilities. Selection of valleys based on topography can reduce the construction cost per ha. (3) Because the internal rate of return of a project depends essentially on costs and benefits, a great deal of attention should be given to these two factors in project preparation documents. (4) In the attempt to reduce project costs, expectations were set too high regarding the population's contribution and ability to render exhausting construction services. (5) Unforeseen problems emerged especially in the early implementation phase of the project.