Small-scale irrigation projects often face serious programming problems such as escalating recurrent costs, inflexible design, nonfunctioning irrigation groups, inadequate agency capacity, and, most critically and ambiguously of all, the problem of determining the optimal mix of state and local responsibilities. Defining small-scale irrigation projects to include those that are managed either by the state or by local groups, this study analyzes key program choices regarding investment, design, local organization, and implementing agency and in so doing develops an initial strategy for overcoming existing problems. Major elements of the strategy include: (1) supporting local investment initiatives; (2) incorporating local input into project design; (3) forming local irrigation groups prior to system construction and giving them autonomy, attempting to understand the functions of existing local groups and the local structure of the ownership and distribution of land and water rights, and training agency staff to work with local groups; and (4) selecting the implementing agency (or agencies) on the basis of specific project needs, assigning this agency a service rather than a managerial role, and developing a support network which includes research and management institutions, other government agencies, and private voluntary organizations. The advantages of the proposed strategy are enumerated in a final section. A 40-item bibliography (1979-84) is appended.