Summarizes attached final evaluation (XD-ABG-919-A) of the command water management (CWM) component of the Irrigation Systems Management Project in India. The CWM component was to provide the basis for relevant agencies to cooperate with water users in better matching available irrigation water and cropping patterns, and in providing necessary materials and services to increase agricultural productivity. Evaluation covered the period from 6/83-6/93. Major findings are as follows. (1) Civil works received the most emphasis and achieved the best results; most project output indicators concerned civil works targets and most funds were budgeted for civil works. Institutional development received much less attention and was hindered by a lack of clear direction. (2) The Subproject Management Office (SMO) was not found to be an effective model of decentralized management. (3) The concept of expanding the Water Users Association (WUA) model to other activities beyond water delivery was flawed. (4) Some agricultural development and water management activities (e.g., land leveling, more appropriate crop and crop variety selection, improved methods for sowing and fertilizer application) helped some farmers achieve higher net incomes and may be replicable on a wider scale. Recommendations focus on civil works; agricultural production and water management; and institutional development. There are also recommendations on the project design, which the evaluators found to be a major shortcoming. The following are among most important recommendations. (1) The concept of a two-stage development, in which civil works must be essentially complete before other important agricultural or institutional activities are started, should be categorically rejected in future projects. Civil and non-civil works must proceed in tandem, and civil works should not be used as an incentive for accomplishing other project activities. (2) Any future CWM-type project should operate through decentralized management having full administrative and financial powers; institutional goals and objectives must be clearly defined at the design stage and objectively verifiable indicators to measure progress toward their attainment must also be specified and monitored over the life of the project. (3) Future activities in agricultural development and water management should be more appropriate to prevailing farming systems, should place greater emphasis on low-cost extension programs, and should promote greater use of high yielding crop varieties. (Author abstract)