mutually exclusive. A recent experimental study challenged this hypothesis and demonstrated that birds are able to detect predators while pecking seeds on the ground (head-down vigilance). Experimental obstruction of head-down vigilance makes birds increase head-up vigilance (i.e. the classical overt vigilance posture). For many foragers in the wild, visibility varies between habitats and foraging methods. We compared the vigilance of Teal Anas crecca and Shoveler Anas clypeata when foraging with their eyes above the water surface (shallow feeding, only the bill submerged) and when foraging with their eyes underwater (deep feeding, head and neck underwater, or upending), at three wintering sites in western France. Birds of both species spent less time in head-up vigilance during shallow foraging than during deep foraging, with no significant difference between sites, which suggests that they are capable of some vigilance during shallow foraging. During deep foraging, the time spent vigilant increased because the frequency of scans was much higher, while scan length decreased. However, these differences could have resulted from variations in the availability of food at different depths. In an experiment where the food availability was constant, we observed the same pattern, with a higher frequency of scans during deep foraging. This study therefore provides strong support for the idea that vigilance and foraging are not always mutually exclusive and shows that switching between searching methods can cause vigilance time – and, as a consequence, loss of feeding time – to vary. This should be taken into account in future field and experimental studies of the trade-off animals make between vigilance and feeding.