Proximity to water is appreciated by households. Hedonic analyses that try to measure the value of this amenity are potentially biased by omitted variables because locations close to water may be selected by households with higher incomes who construct more luxury houses. Because it is difficult to observe all the relevant characteristics, the coefficient for proximity to water may be biased upwards. We circumvent this problem by exploiting a specific characteristic of the Dutch system of planned residential development: often a number of exactly similar houses are constructed close to each other. By comparing the values of such identical houses, we can measure the effect of proximity to water under almost ideal circumstances. Introduction of these fixed effects lowers the estimated impact of this amenity, thereby confirming the conjectured presence of an omitted variable bias.