Regulation and political opposition often force water utilities to rely on nonprice approaches to manage water demand. Using randomized field experiments in three different water utilities, we assess the effectiveness of social comparisons to reduce demand and analyze their interaction with existing conservation programs. In two utilities, the program decreases consumption by 5%, with significant heterogeneity across the distribution of baseline water use. We do not detect a statistically significant average treatment effect in the third utility. Social norms do not appear to crowd out existing conservation programs: treated households are more likely to participate in additional programs. Of the two utilities with significant treatment effects, higher participation rates in conservation programs account for a very small fraction of water savings (3%) in one utility and a modest fraction (9%–25%) in the second. We discuss evidence that social norms may induce participation among the specific type of consumers that utilities wish to target.