The acoustic environments of small, central urban parks are often dominated by traffic sounds. Water sounds can be used to mitigate the negative impacts of unwanted sounds through masking. Studies comparing the effects of different water sounds are typically conducted using recordings in laboratory settings where ecological validity is limited. An urban redesign project in Montreal took the innovative approach of trying three sequential temporary designs of a new public square, each of which included a distinct water feature that produced (...) a lightly-audible mist. Here we report on a field experiment evaluating the effect of the water feature in each of the three designs. Respondents evaluated their experience with the three different designs using questionnaires including soundscape and restorativeness scales, and perceived loudness. The results indicate a significant interaction effect between the water feature and the design of the space, particularly on ratings of chaotic and loud. While two water feature designs had an overall “positive” effect on soundscape assessment, the third water feature design produced the opposite effect. These findings hold even after accounting for ambient temperature. This opportunity to test multiple water features in the same space revealed that water features do not automatically improve soundscape assessments. The visual design, function of the space and environmental conditions should be carefully considered and calls for more field studies. We discuss consequences and considerations for the use of water features in public spaces as well as the implications in terms of ecological validity of soundscape studies. (shrink)