I am interested in the evolutionary history of fishes, and in the genetic tools used in deducing this evolutionary history. DNA sequence analysis-that is, analyzing differences among species and higher groups of organisms in their sequences of base pairs in the DNA molecule in particular regions-has become the paradigm for phylogenetic studies of fishes, amphibians, and reptiles. There are always controversies in science, and I happen to be involved in one of them, having to do with how to use DNA in phylogenetic studies. Phylogeneticists have used both mitochondrial DNA (mtDNA) and nuclear DNA for these studies. However, mtDNA and nuclear DNA have quite different properties, and I have been arguing in the literature that some of the properties of mtDNA violate prerequisites for using them in phylogenetic studies. In a nutshell, the problem is that mitochondria are genetically reduced, once-upon-a-time bacterial intracellular symbionts whose genetic material is extrinsic to the host organism’s genome. This makes mtDNA a questionable proxy for the genome of the host. mtDNA is inherited as a unit and can be subject to intact lateral transfer from one species to another via hybridization, yielding results in conflict with studies basing the phylogeny of the hosts on nuclear DNA. There are valid uses of mtDNA, for example, in the study of the direction of hybridization among species. However, mtDNA should not be used as a primary database in phylogenetic studies. All previous phylogenetic applications of mtDNA should be checked with nuclear DNA sequences and/or other intrinsic data.