There has been broad acceptance among evolutionary biologists of the Ecdysozoa hypothesis that, based principally on molecular phylogenetic studies of small and large subunit ribosomal RNA sequences, postulates a close relationship between molting taxa such as arthropods and nematodes. On the other hand, recent studies of as many as 100 additional genes do not support the Ecdysozoa hypothesis and instead favor the older Coelomata hypothesis that groups the coelomate arthropods with the coelomate vertebrates to the exclusion of the nematodes. Here, exploiting completely sequenced genomes, we examined this question using cladistic analyses of the phylogenetic distribution of 1712 orthologous genes and 2906 protein domain combinations; we found stronger support for the Coelomata hypothesis than for the Ecdysozoa hypothesis. However, although arrived at by considering very large data sets, we show that this conclusion is unreliable, biased toward grouping arthropods with chordates by systematic high rate of character loss in the nematode. When we addressed this problem, we found slightly more support for Ecdysozoa than for Coelomata. Our identification of this systematic bias even when using entire genomes has important implications for future phylogenetic studies. We conclude that the results from the intensively sampled ribosomal RNA genes supporting the Ecdysozoa hypothesis provide the most credible current estimates of metazoan phylogeny.