Dissection of a series of individuals of several species in the avian family Tyrannidae revealed extensive individual variation in a hind limb muscle, M. flexor cruris lateralis (FCRL), especially in the kingbirds. Three possible explanations are explored: (1) the adaptive variation hypothesis, that morphological variation is maintained by selection as a means of subdividing the niche of a population and reducing interindividual competition; (2) the transformation hypothesis, that variation is not maintained but represents a transient stage in the active process of structural reduction and loss; (3) the random variation hypothesis, that there is no selection. A correlation between morphology and behaviour exists in tyrannid species, but was not found in an Old World ecological counterpart, Muscicapa striata. This may reflect divergent phylogeny rather than the unimportance of natural selection in shaping morphology of FCRL. Asymmetry, atavisms, atrophy, heritability of morphology and sexual dimorphism are discussed.