It is conventional wisdom that poor households use less water than rich households, and intuition suggests that an increasing block tariff with a lifeline block will target subsidies to poor households. In this paper we provide a simple diagnostic tool that a water utility can use to estimate the distribution of subsidies to households in different income quintiles and to check whether this intuition about the incidence of subsidies is correct in a specific local service area. The results of our illustrative simulations calibrated using data from low- and medium-income countries, show that subsidies delivered through the most common tariff structures are very poorly targeted to poor households. This finding holds regardless of the specific characteristics of the tariff structure used to calculate households’ water bills. We also find that the higher the correlation between household income and water use, the lower the proportion of total subsidies received by poor households.