Evidence suggests that each cuckoo species is specific to a discrete group of primary hosts, although there is often considerable overlap in the use of secondary hosts. Presumably therefore, the host segregation of cuckoo species has evolved as a result of interspecific competition, in line with Gause's hypothesis. Teleological evidence of host specificity on cuckoos is their diverse egg morphology, which could have arisen either in response to varying degrees of tolerance shown by different host species or as the result of interspecific and intraspecific egg replacement by female cuckoos. We suggest (a) that the second alternative provides the more convincing explanation with regard to host specialization in cuckoos and (b) that host defences in these essentially one-on-one host/parasite systems may operate as a density-dependent control on cuckoo numbers rather than as a continuing influence on the morphology of the cuckoo egg.