There has been recent discussion about the evolutionary pressures underlying the long necks of extant giraffes and extinct sauropod dinosaurs. Here we summarise these debates and place them in a wider taxonomic context. We consider the evolution of long necks across a wide range of (both living and extinct) taxa and ask whether there has been a common selective factor or whether each case has a separate explanation. We conclude that in most cases long necks can be explained in terms of foraging requirements, and that alternative explanations in terms of sexual selection, thermoregulation and predation pressure are not as well supported. Specifically, in giraffe, tortoises, and perhaps sauropods there is likely to have been selection for high browsing. It the last case there may also have been selection for reaching otherwise inaccessible aquatic plants or for increasing the energetic efficiency of low browsing. For camels, wading birds and ratites, original selection was likely for increased leg length, with correlated selection for a longer neck to allow feeding and drinking at or near substrate level. For fish-eating long-necked birds and plesiosaurs a small head at the end of a long neck allows fast acceleration of the mouth to allow capture of elusive prey. A swan's long neck allows access to benthic vegetation, for vultures the long neck allows reaching deep into a carcass. Geese may be an unusual case where anti-predator vigilance is important, but so may be energetically efficient low browsing. The one group for which we feel unable to draw firm conclusions are the pterosaurs, this is in keeping with the current uncertainty about the biology of this group. Despite foraging emerging as a dominant theme in selection for long necks, for almost every taxonomic group we have identified useful empirical work that would increase understanding of the selective costs and benefits of a long neck.