The goals of this research were to 1) identify statistically significant, observable consumer characteristics that are factors of demand for water, 2) econometrically model water demand incorporating these characteristics into the model, which allows us to segment consumer groups, and 3) design a conservation incentive program that allows individuals to choose their own best conservation alternative while, in aggregate, achieving the conservation program goals. We employed an experimental game that simulates water consumption from a potentially exhaustible source. Experiment participants included students at the University of New Mexico and members of communities in New Mexico. We found heterogeneous demand for water, including differences between student and community participants and between those who are employed and those who are retired. Consumption is a function of a variety of social and cultural factors including age, gender, ethnicity, political affiliation, religious affiliation, and risk preferences. Policy based on demand estimates that assume homogenous consumers is not efficient. By disaggregating demand, we show that a menu of price systems can achieve conservation goals with less loss of consumer welfare than can alternative policies. These systems provide each consumer an incentive to choose the one that is most beneficial given his or her unique demand, minimizing enforcement costs.