When policies incentivize voluntary activities that also take place in the absence of the incentive, it is critical to identify the additionality of the policy--that is, the degree to which the policy results in actions that would not have occurred otherwise. Rebate programs have become a common conservation policy tool for local municipalities seeking to retrofit residential properties with efficient appliances. This research evaluates whether such rebates can be cost-effective means for water utilities to promote water conservation. A unique database is developed that combines water-use data over a three-year period for all households that participated in a utility's high-efficiency toilet (HET) rebate program, water-use data for a matched sample of neighbors who did not receive a rebate, and a survey of rebate participants. Difference-in-differences models indicate that installation of an HET reduces household water consumption by approximately 7 percent. While installation of an HET appears to be an effective means for achieving household reductions in water consumption, our results also suggest that the rebate program is a much less effective means for achieving household reductions in water consumption. Specifically, the rebate program is found to provide limited additional water savings beyond what would have occurred naturally and is responsible for only 37 percent of the total water reduction attributable to the installation of HETs over the study period. (Contains 26 footnotes, 7 tables, and 2 figures.)