To test the hypothesis put forward by Feduccia of the origin of modern birds from transitional birds, we sequenced the first two complete mitochondrial genomes of shorebirds (ruddy turnstone and blackish oystercatcher) and compared their sequences with those of already published avian genomes. When corrected for rate heterogeneity across sites and non–homogeneous nucleotide compositions among lineages in maximum likelihood (ML), the optimal tree places palaeognath birds as sister to the neognaths including shorebirds. This optimal topology is a re–rooting of recently published ordinal–level avian trees derived from mitochondrial sequences. Using a penalized likelihood (PL) rate–smoothing process in conjunction with dates estimated from fossils, we show that the basal splits in the bird tree are much older than the Cretaceous–Tertiary (K–T) boundary, reinforcing previous molecular studies that rejected the derivation of modern birds from transitional shorebirds. Our mean estimate for the origin of modern birds at about 123 million years ago (Myr ago) is quite close to recent estimates using both nuclear and mitochondrial genes, and supports theories of continental break–up as a driving force in avian diversification. Not only did many modern orders of birds originate well before the K–T boundary, but the radiation of major clades occurred over an extended period of at least 40 Myr ago, thus also falsifying Feduccia's rapid radiation scenario following a K–T bottleneck.