The phylogenetic relationships among all but one horned lizard species were inferred from mitochondrial ribosomal rRNA gene sequences (251 bp of 12S and 457 bp of 16S) and morphology (32 informative characters). Phylogenies were reconstructed based on separate and combined analyses using parsimony and maximum-likelihood methods. The separate mtDNA and morphological hypotheses were largely incongruent. Bootstrap analyses suggested that most of the incongruence was the result of weak support (bootstraps <70%) for alternative relationships (e.g., placement of Phrynosoma asio). However, there were strongly supported alternative placements of Phrynosoma ditmarsi. Mitochondrial DNA strongly placed P. ditmarsi with the short-horned lizards (i.e., Phrynosoma orbiculare and Phrynosoma douglasi sensu lato), whereas morphology strongly supported its exclusion from the large clade containing the short-horned lizards and placed P. ditmarsi with the southern Phrynosoma braconnieri Phrynosoma taurus. Various consensus methods for assessing topological congruence (= taxonomic congruence) between the separate phylogenies indicate little congruence (i.e., strict consensus) or misleading congruence (i.e., Adam's consensus). Also, although bootstrap analyses suggested strong incongruence only involving the placement of P. ditmarsi, the Wilcoxon signed-ranks test suggests the incongruence between datasets is significant, even when P. ditmarsi is pruned from the trees. Combined analysis of the mtDNA and morphological data resulted in a single most-parsimonious tree. All resolved clades, except two, were also discovered in the separate analyses of mtDNA or morphology.