AbstractKnowledge of the evolutionary history of crown group birds (Neornithes) has significantly improved through emerging congruence among phylogenetic hypotheses and the description of numerous newPalaeogene stem group representatives. However, controversies still persist about the precise interrelationships of many extant and fossil taxa and about the timing of the diversification of the neornithine crown group. Using the example ofPhaethontiformes (tropicbirds) andPsittaciformes (parrots), it is shown how new sequence‐based phylogenies may shed light on the relationships of fossils with an unexpected character mosaic, and how such fossils can improve our understanding of character evolution in morphologically disparate avian taxa. The earliest occurrences of neornithine birds are plotted on a current phylogeny. As noted by previous authors, an extensive diversification of neornithine birds before the latestCretaceous is not supported by the fossil record, and the existence of essentially modern‐type representatives ofTelluraves (the clade including most arboreal birds) in theCretaceous, such as suggested from molecular calibrations, is highly unlikely.