Markets for water rights have long been advocated to resolve user conflicts and to improve efficiency (Trelease 1961; Hartman and Seastone 1970). The potential for efficiency gains from market allocation of water has been well established (Burness and Quirk 1979; Eheart, Lyon, and Wong 1983; Moore 1986). Almost all studies of water markets address the transferability of long-term water rights, i.e., or property rights (Eheart et al. 1983; Lyon 1986; Saliba, Bush, and Martin 1987; Crouter 1987).' However, rental or spot markets for water can improve efficiency by allowing allocational adjustments in line with changing conditions without disturbing the base rights system. Exchanges of either base rights or spot rights that entail a change in the point of diversion may result in third-party effects even in the absence of return flow problems (Johnson, Gisser, and Werner 1981; Eheart et al. 1983).2 Such third-party effects place practical limits on the geographic scope of water exchanges and result in a market, i.e., a market with few eligible participants. Thin markets are likely to be manipulated by participants, leading to inefficiency. Unless suitable bargaining rules are designed to counter market manipulation, the efficiency gains from water exchanges cannot be fully realized. This paper uses multilateral bargaining models based on a game theoretic approach to address bargaining rules that will facilitate the efficient operation of a thin water market across a variety of bargaining environments. Institutional features such as the size of the water market, the water rights system, the distribution of farm sizes, and the level of participants' information define the bargaining environment. The models are translated into computer simulations based on data representing irrigation water demand in the Crane Creek Watershed of Kankakee County, Illinois. Thin water markets are likely to be the rule rather than the exception, especially in the U.S. West where many streams with insufficient flows are fully or overappropriated. Given this fact, it is crucial to investigate how the efficiency of a thin water market is affected by the relative bargaining strengths of the participants and to suggest suitable bargaining rules that could improve the social gains achieved in thin water markets.