Birds differ widely in the shape and size of their tails, yet until re
cently nearly all attempts to understand this morphological diversity
have focused exclusively on the role of sexual selection for elaborate
tail ornaments. In contrast, here we use aerodynamic models and compa
rative data to examine the extent to which natural selection can expla
in observed variation in avian tail morphology. A tail reduces a bird'
s overall lift-to-drag ratio (L/D), but the tail is important in maint
aining stability over a range of flight speeds and in generating lift
to help with turning and slow flight. As predicted, we found that larg
e, open-country birds selected for high LID have relatively short tail
s. Conversely, birds that need high maneuverability in order to feed a
erially or avoid collisions in cluttered environments have longer tail
s. One prediction not supported by our data is that tails should also
be elongated in species that need supplementary lift in order to wind-
hover. Tail shape can also be important, particularly in aerial hawker
s that rely on high agility to obtain their food. We found that, as ex
pected, such birds generally have long, deeply forked tails capable of
generating maximum turning moments for a given drag. Understanding th
e various ways in which natural selection acts on tail morphology prov
ides an essential baseline from which to investigate how sexual select
ion has gone on to modify further the shapes of birds' tails.