The reconstruction of complete mitochondrial genomes (mitogenomes) has considerable potential to clarify species relationships in cases where morphological analysis and DNA sequencing of individual genes are inconclusive. However, the trend to use only mitogenomes for the phylogenies presented in mitogenome announcements carries the inherent risk that the study species’ taxonomy is incorrect because no mitogenomes have yet been reconstructed for its sister species. Here, I illustrate this problem using the mitogenomes of two seahorses, Hippocampus capensis and H. queenslandicus. Both specimens used for mitogenome reconstruction originated from traditional Chinese medicine markets rather than native habitats. Although mitogenome phylogenies placed these specimens correctly among the seahorses from which mitogenomes were available at the time, incorporating single-marker sequence from closely related species into the phylogenies revealed that both mitogenomes are problematic. The mitogenome of the endemic South African H. capensis did not cluster among single-marker DNA sequences of seahorses from the species’ native habitat, but among sequences submitted under the names H. casscsio, H. fuscus and H. kuda that originated from all over the Indo-Pacific, including China. Phylogenetic placement of the mitogenome of H. queenslandicus within a cluster of seahorses that also included H. spinosissimus confirms an earlier finding that H. queenslandicus is a synonym of H. spinosissimus, a widespread Indo-Pacific species that also occurs in China. It is recommended that mitogenome announcements incorporate available single-marker sequences of closely related species, not only mitogenomes. The reconstruction of mitogenomes can exacerbate taxonomic confusion if existing information is ignored.