To prevent drownings in public swimming pools (PSP), French legislation requires constant surveillance by state-certified lifeguards. While previous research showed that surveillance was not always effective, this article focuses on efficiency of surveillance when it is effective. We evaluated the ability of 4 volunteer professional lifeguards to detect a submerged manikin under controlled conditions. One hundred and eight (108) tests were carried out in 2 PSP. Four variables were controlled (i.e., distance, depth, surveillance station, attendance). Our results showed that rapid drowning detection was not exclusively linked to the individual detection capabilities of a lifeguard, but rather it emerged from a tight coupling between the lifeguard’s perception and his/her working environment. The tests performed in this study are useful to prevent drownings and therefore should be reproduced in other public swimming pools (i) to identify problematic situations and (ii) to train lifeguards’ surveillance capabilities.