FLUCTUATING asymmetry (FA) refers to the random, stress-induced deviat
ions from perfect symmetry that develop during the growth of bilateral
ly symmetrical traits(1,2). Individual differences in the level of FA
may influence mate choice(3): in a number of species, females prefer t
o mate with males that have more symmetrical sexual ornaments(4-7). As
the degree of FA has been shown to reflect the ability of individuals
to cope with a wide variety of environmental stresses(2,8,9), it has
been suggested that mating preferences for symmetry evolve for adaptiv
e reasons, because the degree of FA provides honest information about
male quality(10,11). Here I use simple, artificial neural networks to
show that such preferences are likely to arise in the absence of any l
ink between symmetry and quality, as a by-product of selection for mat
e recognition.