Recently, California has made large investments in wastewater recycling to replace applications that use potable water. It may be expected that the use of recycled water reduces potable water use, but such an equivalency is not assured. The addition of recycled water infrastructure in a large Californian water district creates a natural experiment where this work tests how recycled water usage affects primary potable water. This is done using econometric methods for causal inference in an observational setting that mirror a randomized control trial (RCT). From 2001 to 2014, a number of public parks were given recycled water infrastructure, while others in those regions remained exclusively on potable supply. A two-way fixed effects regression is used to produce a difference-in-differences estimate of the average treatment effect of recycled water on total and potable water usage. The results indicate that potable water usage is reduced significantly when a park is connected to the recycled water supply. The estimated rate of displacement in the study period is 81.7%, meaning each unit of recycled water use avoided 0.817 units of potable water usage, which implies the connection of parks to recycled water supply increases total water use. The analysis provides, to the best of our knowledge, the first empirical estimate of the water savings claimed by urban water recycling programs, and the first empirical estimate of displacement using quasi-experimental methods. The methodology can be utilized to evaluate the effectiveness of recycling programs around the world. All rights reserved, Elsevier.