A management model is presented for the optimal development of marginal water sources in arid zones in conjunction with minimizing the dependence on high-quality water. These marginal sources, which may include among others saline groundwater, treated wastewater, and runoff water, are required to augment a limited supply from regional sources. The objective is to minimize operational and capital costs while simultaneously allocating a conventional regional supply in a best way among a set of local sites. A novel aspect is the consideration of water quality as an additional constraint in the decision model. In this way an optimal investment strategy for marginal water source development and use is obtained while satisfying quality requirements at the individual sites. The model formulated takes the form of a mixed binary integer linear problem. The main purpose of the presented model is to delineate a methodology for marginal water considerations and development in arid zones. Water qualities, supply and demand for diverse uses, and related costs are of primary importance. Several simplifying assumptions are made, such as aggregation on an annual basis, in order to cope with the essential features of the problem at a reasonable level of complexity. These assumptions may be relaxed at a later, more detailed stage of analysis. A case of study of the Negev Desert in southern Isreal is shown. An important conclusion is that saline groundwater production at local demand sites should be increased dramatically. The usefulness of sensitivity analysis in the decision process is also demonstrated. ¿ American Geophysical Union 1993