In order to provide an adequate potable water supply for every individual in developing countries by 1990, greatly increased efforts have been slated by donor agencies during the International Drinking Water Decade (1980-90). To be effective, however, these increased efforts must improve on previous water projects, which have had a high failure rate. This paper examines problems associated with rural water supply programs and recommends changes in policy directions. The author briefly reviews the history of water assistance projects and the priority accorded them by donors and host countries and lists recent evaluations conducted by international and bilateral donors. The many maintenance-related problems which are most commonly responsible for failures in water supply projects are discussed. In general, the lack of spare parts and other supplies, coupled with insufficient national program support (often reinforced by donor emphasis on construction, rather than maintenance), leads to equipment breakdown or misuse, and not the technology itself. Lack of coordination among the myriad of donors and consultants leads to further confusion and inefficiency. Lastly, villagers are often unaware or misinformed about the benefits of potable water -- they tend to value taste or convenience over safety. Project planners often ignore the social/religious value placed on water by the community, which can lead to a lack of local motivation to maintain and use the new systems. To ensure the effectiveness of future programs, the author recommends that A.I.D.: (1) choose appropriate technology which is cost-effective and easily maintained; (2) increase donor collaboration; (3) improve the manpower, administration, and commitment of host governments to national water programs; and (4) increase community participation and education related to local water projects. The author especially recommends a shift in emphasis from the planning of individual projects to a national or sector approach in which institutional building activities as well as hardware are funded. A list of practical steps to achieving these policy shifts is presented. A 14-item bibliography (1975-78) and an appendix on problem dimensions are attached.